Subtractive 3 color: Chromogenic monopack
“The Eastman Colour Films are multilayer films of the type in which the layers are not separated after exposure. Films of this class are known as Multilayer, Monopack or Integral Tripack. “Multilayer” is descriptive not only of this particular group of films, but also those in which the layers may be separated after exposure, while “Monopack” is liable to be associated with a particular process which has been quite widely employed by Technicolor. “Integral Tripack” is therefore adopted as the most convenient term for describing the Eastman Colour Films.
Three types of Eastman Colour Film are manufactured. These are the Colour Negative Film, intended for use as the picture negative material in the camera; the Colour lnternegative Film, used for a similar purpose to black-and-white duplicating negative film; and Colour Print Film, which may be employed in preparing prints from either the Colour Negative or Colour lnternegative. A special black-and-white Separation Positive Film is also provided and this is intended for use in preparing three separation positives from the Colour Negative. The separation positives form an intermediate link with the Colour Negative when making a Colour Internegative, so that their function is similar to that of a master positive in a black-and-white system.
Integral tripack camera films have the advantage that they may be used in a standard black-and-white camera, and apart from a check on the colour correction and focus of the lens, no special precautions are necessary. It is of interest to note that the colours of integral tripack negatives, as well as the densities, are reversed compared with the original scene.
The coloured images in Eastman Colour Films are produced by a method known as dye-coupling development. For this a special developing agent is used in conjunction with a second compound known as the colourforming coupler. Photographic development is a process of chemical reduction brought about by the developing agent, which is oxidized in proportion to the amount of silver formed. The oxidized developing agent combines with the colour forming coupler to create a dye of appropriate colour, the concentration of which is proportional to the amount of silver in the image. The dye thus formed must be insoluble in water so that the reaction shall be quite local and a dye image of high resolution obtained. The silver image is removed at a later stage of the process.
Three colour-forming couplers provide the appropriate dyes and are incorporated in the relevant emulsion layers.”
(Craig, G.J. (1953): Eastman Colour Films for Professional Motion Picture Work. In: British Kinematography, 22,5, 1953, pp. 146-158.)
Galleries Hide all Galleries ×Open all Galleries ▼
Cinerama samples on Eastman Color Negative from the Kodak Film Samples Collection at the National Science and Media Museum in Bradford.
Credit: National Science and Media Museum Bradford.
Photographs by Josephine Diecke, SNSF project Film Colors. Technologies, Cultures, Institutions and Joëlle Kost, ERC Advanced Grant FilmColors.
Original Technical Papers and Primary Sources
Craig, G.J. (1953): Eastman Colour Films for Professional Motion Picture Work. In: British Kinematography, 22,5, 1953, pp. 146-158.
Foster, Frederick (1953): Eastman Negative-Positive Color Films for Motion Pictures. In: American Cinematographer, 34,7, July 1953, pp. 322-333, 348.
Hanson, Wesley T., Jr. (1950): Color Correction with Colored Couplers. In: Journal of the Optical Society of America, 40,3, 1950, pp. 166–171, on p. 166.
Hanson, Wesley T. (1955): Subtractive Color Photography. Spectral Sensitivities and Masks. In: Journal of the Optical Society of America, 45, 1955, pp. 476-481.
Keene, G. T.; Clifford, J. D. (1962): Commercial Systems for Making 8mm Prints. In: Journal of the Society of Motion Picture and Television Engineers, 71,6, pp. 447–449.
Smits, J.; Corluy, H.; De Kerf, J. (1966): Farbmetrische Analyse fotografischer Farbwiedergabe-Verfahren. In: Die Farbe, 15, 1966, pp. 102-118. (in German)
Young, Freddie (1966): A method of Pre-exposing Color Negative for Subtle Effect. In: American Cinematographer, 47,8, Aug. 1966, p. 537.
Andrew, Dudley (1980): The Post-War Struggle for Colour. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 40-50, on p. 47.
Anonymous (1956): Current Techniques of 35mm Color Film Photography and Printing. In: American Cinematographer, 37,1, January 1956, pp. 26-27 and p. 58.
Anonymous (1980): Colour Problem. In: Sight and Sound, 50, pp. 12–13, on pp. 12–13and on p. 13.
Batistová, Anna (2013): Glorious Agfacolor, Breathtaking Totalvision and Monophonic Sound. Colour and “Scope” in Czechoslovakia. In: Simon Brown, Sarah Street and Liz Watkins (eds.): Color and the Moving Image. History, Theory, Aesthetics, Archive. New York, London: Routledge, pp. 47-55.
Behlmer, Rudy (1964): Technicolor. In: Films in Review, 15,6, pp. 333–351, on pp. 350–351.
Bergala, Alain (1995): La couleur, la Nouvelle Vague et ses maîtres des années cinquante. In: Jacques Aumont (ed.): La Couleur en cinema. Milan: Mazzotta, pp. 126–136, on p. 129and p. 131. (in French)
Beyer, Friedemann; Koshofer, Gert; Krüger, Michael (2010): UFA in Farbe. Technik, Politik und Starkult zwischen 1936 und 1945. München: Collection Rolf Heyne, on p. 54. (in German)
Borde, Raymond (1988): Die Filmarchive und der Farbfilm. Eine Einführung. In: Gert Koshofer: Color. Die Farben des Films. Berlin: Wissenschaftsverl. Volker Spiess, pp. 7–10, on p. 9. (in German)
Bordwell, David; Staiger, Janet; Thompson, Kristin (1985): The Classical Hollywood Cinema. Film Style and Mode of Production to 1960. London: Routledge, on p. 357.
Cubitt, Sean (2014): The Practice of Light. A Genealogy of Technologies from Prints to Pixels. Cambridge, Massachusetts, London: MIT Press, on p. 135.
Everett, Wendy (2007): Mapping Colour. An Introduction to the Theories and Practices of Colour. In: Wendy Everett (ed.): Questions of Colour in Cinema. From Paintbrush to Pixel. Oxford: Peter Lang, pp. 7–38, on p. 22.
Hanson, Wesley T. (1981): The Evolution of Eastman Color Motion Pictures. In: Journal of the Society of Motion Picture and Television Engineers, 90,9, 1981, pp. 791-794.
Happé, Bernard (1984): 80 Years of Colour Cinematography. London: British Kinematograph Sound & Television Society, on p. 16.
Kisner, W. I. (1962): A New Color Negative Film for Better Picture Quality. In: Journal of the Society of Motion Picture and Television Engineers, 71,10, pp. 776–779.
Kisner, W. I. (1962): A Higher Speed Color Print Film. In: Journal of the Society of Motion Picture and Television Engineers, 71,10, pp. 779–781.
Misek, Richard (2007): “Last of the Kodak”. Andrei Tarkovsky’s Struggle with Colour. In: Everett, Wendy (ed.): Questions of Colour in Cinema. From Paintbrush to Pixel. Oxford: Peter Lang, pp. 161–176, on p. 163, on p. 164 and on pp. 165–167.
Neale, Steve (1985): The Beginnings of Technicolor. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 13-23, on p. 20and on pp. 21-22.
Parker, David L. (1973): ‘Blazing Technicolor,’ ‘Stunning Trucolor,’ and ‘Shocking Eastmancolor.’ In: Tom Shales et al. (ed.): The American Film Heritage, The American Film Institute. Washington, D.C.: Acropolis Books Ltd., pp. 19-27.
Penning, Lars (1988): Farbe im klassischen Piratenfilm. In: Karl-Dietmar Möller-Nass, Hasko Schneider and Hans J. Wulff (eds.): 1. Film- und Fernsehwissenschaftliches Kolloquium. Münster: MAkS, pp. 36–40, on p. 38. (in German)
Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on p. 161, on pp. 191–192 , on p. 201 , on pp. 207–208 , on p. 209 and on pp. 216–218. (in Italian)
Pinel, Vincent (1992): La forêt des techniques. In: Michel Ciment (ed.): Ciné mémoire. Colloque international d’information (7-9 octobre 1991). Paris: Femis, pp. 17-24, on pp. 21-24. (in French)
Ryan, Roderick T. (1977): A History of Motion Picture Color Technology. London: Focal Press, on pp. 148-158.
Stokes, Melvyn (2009): Colour in American Cinema. From The Great Train Robbery to Bonnie and Clyde. In: Raphaëlle Costa de Beauregard (ed.): Cinéma et couleur. Paris: M. Houdiard, pp. 184–192, on p. 187.
Webers, Johannes; Westendorp, Kurt (1979): Einführung in die Kopierwerktechnik (XIII). In: Fernseh- und Kinotechnik, 33,6, pp. 213–215. (in German)
“Eastman Colour Films For Professional Motion Picture Work
G. J. Craig, O.B.E. (Fellow)
Read to a meeting of the Film Production Division on April 15, 1953.
In modern processes of colour photography a technique has been established by which colours are analyzed in accordance with their reflectances within the three broad spectral regions, blue, green and red. This can be done, either by taking three separate photographs through blue, green and red filters on normal panchromatic emulsions, or by using emulsions with colour sensitivities localized as far as possible to the appropriate spectral bands.
The photographs may be recorded on three separate films, as in the Technicolor camera, or on a single normal picture negative film, as in the successive frame method used for animation work. Yet another system is one in which a multilayer film is employed, with the emulsions coated one above the other on a single support. If the three layers can be separated after exposure then black and white images may be used — Multilayer Stripping Film1 is an example of this technique. But if the multilayer is integral, and not meant to be separated, the images must be selectively coloured to allow for subsequent separation by colour analysis.
When the three separation records have been made, there are two basic methods whereby the scene can be reproduced in colour: additive and subtractive synthesis. The additive method is outside the scope of this paper.
The subtractive method makes use of certain dyes and pigments which have the property of selectively absorbing blue, green or red light while transmitting the remainder of the spectrum. These dyes and pigments are called the subtractive primaries, and they control the amounts of blue, green and red light in a mixture by absorbing or subtracting from white light the amounts of blue, green and red not required.
For this purpose, three positive images in dye or pigment form are prepared from the separation negatives, the red separation negative being printed in the red-absorbing or Cyan dye; the green negative in the green-absorbing Magenta dye; and the blue negative in the blue-absorbing Yellow dye. The three images may then be superimposed on a single support to make the completed colour reproduction.
The Eastman Colour Films are multilayer films of the type in which the layers are not separated after exposure. Films of this class are known as Multilayer, Monopack or Integral Tripack. “Multilayer” is descriptive not only of this particular group of films, but also those in which the layers may be separated after exposure, while “Monopack” is liable to be associated with a particular process which has been quite widely employed by Technicolor. “Integral Tripack” is therefore adopted as the most convenient term for describing the Eastman Colour Films.
Three types of Eastman Colour Film are manufactured. These are the Colour Negative Film, intended for use as the picture negative material in the camera; the Colour lnternegative Film, used for a similar purpose to black-and-white duplicating negative film; and Colour Print Film, which may be employed in preparing prints from either the Colour Negative or Colour lnternegative. A special black-and-white Separation Positive Film is also provided and this is intended for use in preparing three separation positives from the Colour Negative. The separation positives form an intermediate link with the Colour Negative when making a Colour Internegative, so that their function is similar to that of a master positive in a black-and-white system.
Integral tripack camera films have the advantage that they may be used in a standard black-and-white camera, and apart from a check on the colour correction and focus of the lens, no special precautions are necessary. It is of interest to note that the colours of integral tripack negatives, as well as the densities, are reversed compared with the original scene.
The coloured images in Eastman Colour Films are produced by a method known as dye-coupling development. For this a special developing agent is used in conjunction with a second compound known as the colourforming coupler. Photographic development is a process of chemical reduction brought about by the developing agent, which is oxidized in proportion to the amount of silver formed. The oxidized developing agent combines with the colour forming coupler to create a dye of appropriate colour, the concentration of which is proportional to the amount of silver in the image. The dye thus formed must be insoluble in water so that the reaction shall be quite local and a dye image of high resolution obtained. The silver image is removed at a later stage of the process.
Three colour-forming couplers provide the appropriate dyes and are incorporated in the relevant emulsion layers.
Colour Rendering and Image Sharpness in Integral Tripack Processes
At present, there are two basic problems associated with the design of a satisfactory integral tripack process. The first of these is accuracy of colour rendering which, of course, is not confined to the integral tripack system but is of special importance in this case because of the practical problems involved in correcting deficiencies. The second factor is image sharpness, necessarily important because the optical image loses sharpness by light scatter when it passes through successive emulsion layers. Special measures have been adopted to meet these problems in the design of Eastman Colour Films.
Certain theoretical difficulties in the accuracy of colour rendering by additive and subtractive processes have already been described before this Society.2 Another practical problem is concerned with the transmission characteristics of the dyes formed in colour development. Ideally, the dyes used in a subtractive process should completely absorb one-third of the visible spectrum and completely transmit two-thirds. In practice, the available dyes fall considerably below this aim. Consequently there is appreciable colour degradation, even when only one set of dyes is involved as in a reversal process, if the original camera film is subsequently used as the projection positive. Degradation is considerably greater when a negative has to be printed on a positive material having dyes with similar characteristics, and worse still after passing through a duplicating stage. Transmission characteristics of ideal and practical subtractive dyes are shown in Fig. 1.
Automatic Masking by Coloured Dye Couplers
In Eastman Colour Films the transmission characteristics of the yellow dye are fairly good, but the magenta dye absorbs rather heavily in the blue where there should be full transmission, the degree of absorption varying with the dye strength. Similarly the cyan dye absorbs in the blue and green, where it should fully transmit. These shortcomings, if uncorrected, would have considerable adverse effect on the colour quality, and the reproduction would be desaturated and false in colour rendering.
This situation is a familiar one in subtractive colour processes, and some correction is often attempted by a technique known as masking, in which one or more compensating weak positive images are combined with the negative during printing. Such masks are used with the intention of cancelling out the unwanted blue and green absorptions in the cyan and magenta dyes of the negative image. Eastman Colour Negative and Internegative Films incorporate a very valuable masking device, the coloured dye-coupler, which provides the necessary correction automatically.
It will be appreciated that if the magenta dye is absorbing some blue, and the cyan dye some blue and green, then the final print will appear desaturated and wrongly coloured. If, however, the unwanted blue and green absorptions can be made constant, no matter what the dye image density, it will be easy to rectify the deficiency during printing, simply by placing a suitable colour compensating filter in the printing light beam, or alternatively by adjusting the overall colour balance of the tripack print film.
The coloured dye-couplers used for the formation of the magenta and cyan dyes in the appropriate layers achieve this in the following manner. The magenta dye-coupler is coloured yellow and absorbs in the blue region of the spectrum where unwanted absorption is taking place. The coupler is used up as the dye image is formed, so that where image density increases, the colour strength of the coupler decreases. Thus where there is high image density, and therefore high unwanted blue absorption, there is low coupler concentration with consequent low blue absorption. The coloured coupler can be selected so that its maximum absorption corresponds to the blue absorption of the magenta dye at maximum density. In such a case the combined blue absorption of the magenta dye and residual coloured coupler is constant, and remains substantially so at all intermediate densities. A similar function is performed by the coloured coupler in the cyan layer, which is reddish-orange in colour. The behaviour of these coloured couplers is represented diagrammatically in Fig. 2.
The colours of these two couplers result in an orange hue which is visible in the processed negative and obscures the actual colours in the negative to an extent which makes visual assessment impossible. This system of masking is adopted only in the negative stages of the colour process: it cannot be embodied in the final print because of the orange hue of the couplers.
When considering maximum sharpness in a subtractive type colour process it must be remembered that the magenta image has the greatest influence upon the sharpness of a projected picture, the cyan image is next in importance and the yellow image least. It is therefore desirable in an integral tripack film, because of the light scatter in the layers during exposure, to arrange for the magenta image to appear in the top layer, the cyan image in the middle and the yellow image at the bottom.
With the camera negative film such an ideal arrangement is not possible, for in practice the blue sensitive layer (forming the yellow image) has to be placed at the top. A high speed picture negative emulsion is strongly responsive to blue light, so that although two of the three layers in the film can be selectively sensitized to green and red, the fundamental blue sensitivity of the silver halide emulsion is always present and must be eliminated from the green and red sensitive layers by the use of a yellow filter. It is for this reason that the blue sensitive layer must be placed at the top; in such a position it is a simple matter to locate beneath it a yellow filter layer which will completely remove the blue light while transmitting green and red. The filter layer can be designed to be eliminated during processing of the film.
In the print film it is possible to achieve the desired dye layer arrangement because high emulsion speed is not essential, as in the camera negative material, and this permits selection from a wider range of emulsion types.
The internegative film presents a special problem, for although it is a low speed material, and as such might follow the same design as the print film with regard to layer positioning, the fact that coloured couplers are incorporated for automatic masking complicates matters. The blue-absorbing yellow and orange-red couplers would prevent an image being registered on the blue sensitive emulsion if it were at the bottom. The problem is solved by adopting a noncomplementary relationship between the dye image colour and spectral sensitivity of each layer.
Non-Complementary Dye Relationship
In conventional integral tripack materials Yellow, Magenta and Cyan dyes are used for the formation of images originating in the Blue, Green and Red sensitive emulsions respectively. This is because each dye absorbs in the particular spectral region to which the emulsion is sensitized, while freely transmitting (at least, in theory) all other parts of the spectrum. In other words, the dyes bear a complementary relationship to the spectral sensitivities of their respective layers. They control light transmission in exactly the same way as does a silver image, but exercise of this control is confined to light of wavelengths within the colour sensitivity region of the appropriate emulsion. If, therefore, blue, green and red are accepted as the normal regions for which the three emulsions are sensitized, then the normal complementary relationship with the dye image will be Blue/Yellow, Green/Magenta, Red/Cyan. Such a relationship is obligatory, for example, in a reversal type camera film intended for subsequent use in projection; but when integral tripack films are employed at an intermediate stage between the camera negative and final print it is by no means essential for this complementary relationship to be maintained, so long as it can be established in the print. These films may therefore have a non-complementary dye relationship.
Eastman Colour Internegative Film has a non-complementary dye relationship in that the top emulsion layer is blue sensitive and forms a magenta image, the green sensitive middle layer forms a cyan image, and the red sensitive bottom layer a yellow image. In this way, the desired layer positions for the subtractive images are obtained.
Use of this technique in the internegative film is permissible as it is used in conjunction with black-and-white separation positive films. Suitable light filters can be employed when printing from the separation positives on to the internegative to match the noncomplementary dye relationship of the latter. (See Fig. 3).
Preparation of an Internegative
To make an internegative, three separation positives are first prepared. These represent separate records of the three dye images in the original negative. They are obtained by printing the negative successively on the three films using blue, green and red printing lights of suitable quality. Ignoring dye deficiencies, the blue light is obstructed only by the yellow layer in the negative and freely transmitted by the magenta and cyan layers: in the same way green light is affected only by the magenta layer and red light by the cyan layer.
The blue record separation positive is called the Yellow printer, green record the Magenta printer, and red record the Cyan printer, because those are the respective dye layers they control in the internegative.
The separation positives are now successively printed on to the internegative film, suitable light filters again being used to restrict the exposure from each positive to the appropriate emulsion layer of the film. It will be apparent, therefore, that red light must be used with the Yellow printer, green light with the Cyan printer and blue light with the Magenta printer. Production of an internegative in this way is shown schematically in Fig. 3.4
General Properties of Eastman Colour Films
(a) Eastman Colour Negative Film.
Eastman Colour Negative Film is at present manufactured in two types, identified as 5247 and 5248. Type 5247 is suitable for use in daylight illumination, or in high intensity arc light with a straw-coloured filter, such as the Brigham Y.l, before the lamp. In either case a Wratten No. 2B filter should be used on the camera lens to absorb ultra violet radiation. The Exposure Index* for Type 5247 is 16.
Type 5248 is colour balanced for use in tungsten lamp illumination at a colour temperature of 3200°K. In practice, CP lamps operating at 3350°K can also be used, since correction for small differences in colour temperature of the light source can easily be applied during printing. It is not advisable, however, to attempt the lighting of subjects with lamps of mixed colour temperature, at least until ample experience has been gained.
Type 5248 has an Exposure Index of 32 to tungsten illumination. As a guide to lighting level, a key light of 200 foot candles at f/2 is recommended, although lower illumination values can be used if necessary without seriously affecting the quality of the negative. Practical camera tests are always advisable as a check on correct exposure.
The film may be exposed in daylight, or high intensity arc plus a straw-coloured filter such as the Brigham Y.l. In either case a Wratten No. 85 filter is required on the camera lens. In these conditions the Exposure Index becomes 24.
When working with the Colour Negative lighting contrast should be fairly soft, with evenly distributed illumination. Good modelling can be obtained with lighting ratios lower than those used for black and white work.
Although both Type 5247 and Type 5248 are manufactured at present, it is probable that in the near future production will be concentrated on 5248, especially since the two types cannot readily be intercut for printing purposes.
A diagrammatic representation of the emulsion layers in Eastman Colour Negative Film, before and after processing, is shown in Plate 1.
(b) Eastman Colour Print Film.
The type identification number for Eastman Colour Print Film is 5381, while an improved Type 5382, giving increased sharpness, is now being manufactured.
The film is colour balanced for exposure to tungsten lamp illumination. It is appreciably slower than Eastman black-and-white Fine Grain Release Positive Film Type 5302. As a guide to speed, an average density Eastman Colour Negative will print at light 10 on a Bell Howell Model D printer fitted with a 300-watt Reflector Lamphouse and the necessary colour balancing filters in the light beam. A Wratten No. 2B filter must always be used when printing on to the Colour Print Film in order to absorb the ultra violet radiation from the light source, which would otherwise desaturate the colours in the print.
Any standard printer suitable for black-and-white work can be used in making Eastman Colour Prints, provided it has a tungsten lamp light source and means are available for inserting colour balancing filters in the light beam to correct for differences in colour balance of the negative or print film. Such a printer does not permit of the rapid filter changes often required for scene-to-scene colour balance shifts, and for this work special equipment is necessary.
A preferred type of printer is one in which the illumination at the printing aperture is obtained by mixing light from the source after it has passed in three separate beams through narrow band red, green and blue filters, the intensity of each beam being separately controlled.5 A diagram of the optical arrangements for such a printer appears at Fig. 4. Not only does a printer of this type provide a wider range of colour balance changes, but problems associated with the use of colour balancing filters, such as fading and complications due to unwanted absorptions in the filters themselves, are avoided. Moreover, frequent replacement of filter material is liable to become an appreciable cost item. Apart from these advantages tests have shown that improved quality is obtainable on Eastman Colour Print Film exposed in this way.
The sound track on Eastman Colour Print Film is composed of both the silver and dye images, from which very satisfactory sound quality is obtainable.3 A point of interest is that the film is perforated with American Standard Combination Negative-Positive perforations.
A diagrammatic representation of the emulsion layers in Eastman Colour Print Film before and after processing is shown in Plate 1.
(c) Eastman Colour Inter negative Film.
Eastman Colour Internegative Film, designed for use in conjunction with Eastman Colour Negative Type 5247, is identified by the number 5243. A second internegative material, similar in characteristics to 5243 but designed for use with Eastman Colour Negative Type 5248, is known as Type 5245. As has been explained, Colour Internegative has a non-complementary dye relationship and a spectrogram illustrating this is shown at Plate 1. A diagrammatic representation of the emulsion layers before and after processing is also seen in Plate 1.
Colour Internegative Film requires a printer with a high intensity light source in order to obtain adequate exposure. This is partly because the light filters employed for printing the separation positives on to the internegative have narrow transmission bands to restrict exposure to the correct emulsion layer in the film. Additional reasons are the low inherent speed of the internegative film and the necessarily high minimum density of the separation positives. As a guide to the type of printer required, adequate exposure has been obtained from separation positives of 0.8 minimum density in an equipment fitted with a 1000-watt tungsten lamp, reflector, and condenser system, and running at 25feet per minute. A registeringtype printer is essential for printing from separation positives.
(d) Eastman Panchromatic Separation Film, Type 5216.
This material, specially designed for making black-and-white separation positives from colour negative originals, is broadly similar to the black-and-white Eastman Fine Grain Duplicating Negative Film, Type 5203. However, it is capable of a somewhat higher contrast, and the definition is superior to that obtained with 5203. The film has an extended red sensitivity to allow the use of a Wratten No. 70 filter for exposure of the red separation positive. An absorbing dye is incorporated in the emulsion to improve definition, and this is not completely discharged during processing. Consequently, separation positives prepared on this material have a characteristic greenish tint.
Comparative spectrograms for Type 5216 and Type 5203 film are shown in Fig. 5, and typical sensitometric curves for Type 5216 appear in Fig. 6.
Processing of Eastman Colour Films.
Apart from an extra stage to permit removal of the silver image, the processing of Eastman Colour Negative, Colour Print and Colour Internegative Films does not differ fundamentally from the techniques applied to black-and-white materials. But control of solution temperatures and bath compositions has to be much more precise, and the processing machine needs to be built to more rigorous specifications than would be warranted for handling black-and-white films. Discretion is necessary in choice of the materials for machine construction, and efficient squeegee systems are needed at several points, especially when processing Colour Print Film, just before the sound track redevelopment stage. Washing is essential several times during the process, and to minimize reticulation problems it is necessary for the water temperature to be maintained within two or three degrees of the other solutions. Fig. 7 shows in diagrammatic form the various processing stages.
All Eastman Colour Films have a soluble, opaque black backing applied for the purpose of reducing halation, and this must be removed before development. It is first softened in a special bath, and from this the film passes to a rinse tank where the backing is completely removed by buffing the support side of the film, using a roller covered with a soft material such as sponge rubber rotating in the opposite direction to the film travel.
After removal of the backing the film enters the developing solution. It is not greatly different from a black-and-white picture negative developing solution, apart from the special developing agent used, a very low sulphite content, and a higher pH. Development takes 12 to 15 minutes at 70°F., a combined silver and dye image being formed. In black-and-white processing the time can be varied considerably in order to control the ultimate contrast of the negative. Such methods are inadvisable when processing colour film because three emulsions are being developed simultaneously, and they may not all react similarly to a given change in development time. This can lead to an alteration in colour balance.
Development is followed by a brief spray rinse, and the film then passes into the first fixing bath, which is of the normal type used in black-and-white processing. Here the undeveloped silver halide is removed and the emulsion is hardened. A wash follows, after which the film enters the bleach bath where the silver image is converted to silver bromide. After a further wash the film passes to the second fixing bath, identical in composition with the first. Here the silver bromide into which the silver image was converted by the bleach bath is removed, leaving the dye image.
In the case of Colour Print Film, if a sound track has been printed there is a further step which occurs prior to the second fixing bath. A sound track composed of dye image alone is not so satisfactory as one which incorporates the silver image. It is therefore desirable to redevelop the silver bromide in the sound track area only, and this is done by applying a high energy developing solution to the sound track after the wash following the bleach bath. A convenient method of application is by picking up the developing solution on an applicator wheel which forms a bead of developer on the emulsion surface of the film in line with the sound track. In order to prevent spreading of the developing solution into the picture area, development time is kept to a minimum and the solution is made viscous by the inclusion of suitable ingredients. A diagram and photograph of a sound track applicator wheel are shown in Fig. 8.
Following a final wash after the second fixing bath, the film may be passed directly into the drying cabinets if it is Colour Negative or Internegative, or alternatively it can first be run through a tank containing wetting agent to minimize the risk of water spotting. Colour Print Film must be passed through a final stabilizing bath, which, if desired, may also include wetting agent. The stabilizing agent, consisting of formaldehyde, is necessary because the processed print film contains unused colourless colour-forming couplers, which may cause fading of the dye image. These couplers are converted to an inert form by the action of the formaldehyde.
The increasing availability of integral tripack colour films brings nearer the time when all motion picture productions for the commercial entertainment field will be in colour. While the inherent characteristics of the colour film itself must always be a primary factor in determining ultimate quality, it cannot be emphasized too strongly that, far more than in black-and-white photography, good results depend on a high standard of technique in the camera work and film processing. Cameramen accustomed to working with black-and-white films will find adjustments in method essential if good colour quality is to be obtained. Lighting can be less dramatic, exposure must be somewhat more precise, colour temperature of the light source becomes an important factor, and it should always be borne in mind that colours seemingly similar to the eye may not appear so to the colour film. Laboratories must assimilate new methods of sensitometry, and extend their chemical control systems if the requisite standard for colour film processing is to be maintained. Unless these needs are recognized and met, disappointment will surely follow.
The author’s thanks are due to Mr. J. A. Carter of the Kodak Research Laboratories for much helpful advice during the preparation of this paper.
R. H. BOMBACK: I recognized the film projected as a print from a WarnerColor negative. Have you any comments on this?
THE AUTHOR: Eastman Colour Films are used in the WarnerColor process.
W. LASSALLY: You said in your lecture that intercutting of the daylight and tungsten type colour negative was difficult. If this is so, what is the recommended procedure for intercutting exterior and interior scenes?
THE AUTHOR: It is true that intercutting of 5247 and 5248 negatives is difficult on account of the difference in the coloured coupler density. As I explained, future production will probably be concentrated on the 5248 negative, with the intention that it should be used both for exteriors and interiors, with a suitable compensating filter in the case of exterior work.
W. S. BLAND: I notice that the preferred type of printer projects three separate beams through narrow band filters for analysis of the light. These beams pass through different parts of the bulb and so will be subject to their different rates of blackening. Will this not upset the colour balance? In my experience lamps of this type are subject to rather rapid loss in light output through blackening of the bulb? What is the effective life of the lamp?
G. W. ASHTON: Possibly Mr. Bland did not realise from the diagram that the light reaching the printing aperture is automatically controlled to a pre-determined value by the photoelectric cell assembly. Therefore, change in light output due to blackening of the bulb is automatically compensated.
THE AUTHOR: That is true, but the question of the total lamp life must still remain a problem. At the present time this is secondary to the problem of getting sufficient exposure on the film.
R. L. HOULT: You mention that when using 5248 film in daylight a Wratten 85 filter is recommended for colour temperature correction. If 5248 is colour balanced for 3200 K, why is a Wratten 85B filter not recommended for this purpose?
THE AUTHOR: The Wratten 85 filter corrects to one colour temperature only. In practice colour temperature of daylight varies over a wide range, so that a correction at the printing stage is nearly always necessary. In these circumstances it does not much matter whether an 85 or an 85B filter is used. Work is going on with a view to producing a series of colour correction filters intended for use with 5248 film. These will provide more specific colour correction under varying conditions of daylight.
I. B. M. LOMAS: In view of the amount of work which has been carried out on automatic masking methods in recent years, would it not be true to say that ideal subtractive dyes are impossible to obtain?
THE AUTHOR: I think it is reasonable to say that the attainment of ideal dyes in practice seems to be virtually impossible.
F. BUSH: The speaker has said that 5248 film is colour balanced for 3200°, and that satisfactory results are obtainable at 3350°K. Could he say from experience what a practical lower limit to the colour temperature would be, while still maintaining reasonable exposure balance in the three emulsion layers.
THE AUTHOR: I have no practical experience on which to base an answer, but I would estimate it at 3000°K.
A MEMBER: Colour film for daylight use is usually balanced to Mean Noon Sunlight. Is it possible to correct, in the printing process, for daylight conditions other than this, so as to obtain a uniform effective colour temperature in the print?
THE AUTHOR: Certainly, some correction along these lines is possible, but I would not like to suggest that extreme conditions can be corrected in this way.
A MEMBER: Could you say if the Eastman Colour Films are likely to prove more economical than other types of colour processes in current use?
THE AUTHOR: The answer to this depends on the number of prints you wish to have. For quantity production of release prints there are other methods which may well prove cheaper, but as the number of prints required decreases this discrepancy disappears, and may even go the other way.
S. GOOZEE: No mention was made by the speaker of a colour chart to be used when shooting with Eastman Colour Negative so as to provide the laboratory with a check on the actual exposure conditions. Is it intended to provide such a chart?
THE AUTHOR: Yes, a suitable chart will be made available in due course.
B. HONRI: Ealing Studios have used a reproduction of the Ilford colour chart on metal for underwater photography, using Eastman Colour Negative. At 25 feet or more, the chart (and everything in the water surrounding it) appeared greenish blue in the print but as the camera was brought nearer the chart the colour rendering improved until at a distance of five feet perfect colour rendering was obtained.
W. S. BLAND: What width of sound track may be used on Eastman Colour Print Film while still obtaining even density? I had the impression that the bead of developer might be hard to apply over a wide track in order to obtain a uniform density. What sort of dynamic range is obtainable from the Colour Print sound track?
THE AUTHOR: Undoubtedly some care is necessary in designing the applicator for sound track re-development, but no difficulty should be experienced in obtaining uniform re-development on standard tracks. With regard to the dynamic range, I would refer you to the paper “Sound Track on Eastman Color Print Film,” by C. H. Evans and J. F. Finkle, published in the August, 1951, issue of the Journal of the Society of Motion Picture and Television Engineers.
W. LASSALLY: Is any re-focusing of the camera lens necessary if Eastman Colour Negative is used in place of black-and-white negative?
THE AUTHOR: This might be necessary for really critical work.
B. TILL: In my experience, better colour quality is obtainable on colour films at large lens apertures as against small, say f/2.8 as against f/11. Is there any reason why neutral density filters should not be used so as to be able to work at large apertures in bright lighting conditions?
THE AUTHOR: I do not know why you should find better quality at large apertures, but certainly neutral density filters can be used to cut down the light, always remembering that such filters are seldom truly neutral. This fact may lead to printing complications if intercutting is necessary between negatives exposed in various cameras, using an assortment of neutral density filters. If only one neutral density filter is in use, there is no problem.
F. BUSH: With reference to the last question, at large lens aperture there is usually some image flare which lowers the overall contrast and desaturates the colours slightly. Generally speaking, the public seems to prefer this quality to the more saturated higher contrast result obtained at small lens apertures.
P. W. DENNIS: Is there any theoretical or practical reason why coloured couplers as described should not be used in a reversal type film intended as a “master” from which copies are to be made?
THE AUTHOR: I see no reason why coloured couplers should not be used in this way.
P. W. DENNIS: Will WarnerColor negatives made in America be printed on Eastman Colour Print Film in this country?
THE AUTHOR: In America, prints from WarnerColor negatives are made on Eastman Colour Print Film. As this film may not be available in sufficient quantity in this country, it is difficult to say what may happen in this respect. Eastman Colour Negative Film is quite a flexible medium in the sense that other colour printing techniques can be employed apart from Eastman Colour Print Film.
S. IRVING: Why is it that lenses suitable for use in black-and-white photography may not be suitable for Eastman Colour Films?
THE AUTHOR: This depends on the degree of colour correction in a given lens. Black-and-white films, especially if used with light filters, will not require such complete correction in the lens as do integral tripack colour films. A practical test with the particular lens and film is advisable.
A. HINDS: Are Eastman Colour Films easily obtainable?
THE AUTHOR: Supply of the films in Great Britain is subject to a Board of Trade Import Licence. Whether this is granted will depend on the purpose the film is wanted for. So far as actual supplies are concerned, adequate quantities of all types should be available later this year.
MR. WILLIAMS: In order to avoid using internegative, would it be possible to introduce fades and so on by using A and B roll printing technique?
THE AUTHOR: Yes, that should be quite possible. American practice, however, appears to be to make internegatives for all optical inserts.
W. LASSALLY: I thought that, apart from a slightly increased graininess, the internegative sequence we saw in the projected colour print matched extremely well with the sequence from the original negative.
Complete information on Eastman Colour Films is contained in the publication “Production of Motion Pictures in Color using Eastman Color Film.” A copy is available for reference only in the Society’s Library.
1 Capstaff, J. G.: “An experimental 35mm. multi-layer stripping negative film. ” J. Soc. Mot. Pic. & Tel. Eng. , 54, 445, 1950.
2 Hunt, R. W. G.: “Colour cinematography and the human eye.” Brit. Kine. , 19, 173, 1951.
3 Evans, C. H., and Finkle, J. F.: “Sound track on Eastman Color Print Film.” J. Soc. Mot. Pic. & Tel. Eng. , 57, 131, 1951.
4 Anderson, C. R., Groet, N. H., Horton, C. A. and Zwick, D. M.: “An intermediate positive-internegative system for color motion picture photography.” J. Soc. Mot. Pic. & Tel. Eng. . 60, 217, 1953.
5 Streiffert, J. G.: “A fast-acting exposure control system for colour motion picture printing.” J. Soc. Mot. Pic. & Tel. Eng. , 59, 410, 1952.
* This Exposure Index is suitable for use with Weston, General Electric and similar exposure meters with or without the calculators for ASA Exposure Indexes. It may be regarded as an equivalent to’the British Standard Exposure Index (arithmetical). ”
(Craig, G.J. (1953): Eastman Colour Films for Professional Motion Picture Work. In: British Kinematography, 22,5, 1953, pp. 146-158.)
“Eastman Negative-Positive Color Films For Motion Pictures
by Frederick Foster
In recent years, a number of negative and positive color films of the integral tripack type have been made available to the motion picture industry. Their use has been greatly encouraged by the flexibility offered by the negative-positive system, which enables a studio to produce color films with the same ease it does black-and-white. This has been especially true more recently in the production of many three-dimensional films, where 3-D cameras taking single film strips are employed instead of Technicolor 3-strip cameras.
In this respect, the new Eastman color films — three in all: color negative, color positive, and color internegative — have had wide use, and now that the manufacturer has increased the output of these films, their use will become even more general. At present, nearly every major studio in Hollywood is using Eastman color negative in one way or another. Some are using the entire color series. Examples are Warner Brothers, whose Warner-Color system employs Eastman negative and positive film, and Republic Studios whose Trucolor process also employs the full range of Eastman color films. Warner’s House of Wax is an outstanding example of all-Eastman color film use. Twentieth Century-Fox studio is using Eastman color negative in its cameras in the production of CinemaScope films. Columbia Studio uses Eastman color negative in shooting all its 3-D films, with the release prints being produced by Technicolor Corporation.
In all, Eastman Kodak now offers four different film materials which can be used in color productions, such as those outlined above, or which can be used in conjunction with existing commercial color motion picture production processes. Three of these materials represent improvements over earlier Eastman color films which were used in the last few years for a number of motion picture productions.
The most acceptable systems for color motion picture production require the use of intermediate steps in order to include special effects and to provide protection masters. A number of systems are possible when working from a color original, but the preferred system appears to be one employing black-and-white separation positives and an integral tripack-type color internegative, For this, Eastman Kodak has provided special film stocks.
The key film, of course is the negative. The new Eastman Color Negative Film, Type 5248, is balanced for use with tungsten illumination at 3200°K. and requires no filters over lights or lens. It can also be used with daylight or carbon-arc illumination when a Kodak Wratten filter No. 85 is used on the camera lens. The speed is high enough to allow sufficient exposure at a keylight level of about 200 footcandles at F/2.0. It has a tungsten exposure index of about 24, and about 16 for daylight. These are only average specifications and, in many cases, satisfactory exposure can be obtained at even lower lighting levels.
This new film has somewhat lower graininess than the earlier Eastman color negative, and improvements have also been made in the colored couplers to allow better rendition of blue subjects. This results in a lower blue density for the processed film, which is advantageous in printing.
The new Eastman color negative now makes it possible for producers of 35mm films in many fields to make pictures in color using any type of 35mm camera. It is expected that soon we shall see newsreels in color, and more and more explorers and travel and documentary film makers are certain to turn to color, using Eastman color negative in portable Eyemo, Camerette, and Arriflex cameras.
The new Eastman Color Print Film. Type 5382 (35mm) and Type 7382 (16mm) is similar to the earlier product, but improvements have been made to provide better image sharpness. A new magenta coupler is also incorporated in this film which gives better rendition of red hues than was the case with the earlier film.
Printing of the color negative onto Color Print Film can be done with either subtractive type printers employing color compensating filters, or with additive-type printers which utilize three filtered light beams (obtained from three separate sources or from a single source with beam-splitters). In either case, the printer must be designed to permit adjustment of both the intensity and color balance of the light for printing each scene. Additive-type printers have been found to give the best results from the standpoint of good color contrast and saturation.
The sound track can be printed from conventional black-and-white sound negatives prepared in the usual way. Either variable density or variable width tracks may be used. It has been found that better frequency response is obtained if the sound track exposure is confined to the two top emulsion layers of the print film instead of to all three layers.
When effects are to be included in the production, black-and-white separation negatives are made through appropriate filters on Eastman Panchromatic Separation Film, Type 5216, using registering-type printer. Preparation of such separations also provides protection against damage to the valuable color original or against possible fading of the dyes. This step also permits slight corrections for contrast and density variations which have occurred in the exposure and/or processing of the color negative original. The separation positives are processed in a standard black-and-white negative developer by conventional methods.
The separation positives are printed onto a new type Color Internegative Film (Type 5245) using a registering printer. This new film has slightly higher contrast characteristics than the earlier film, hence requires somewhat lower contrast separation positives than was required for the earlier product. As with the Color Negative, improvements have also been made in this film, with the result that there is better rendition of blue subjects. The separate layers of the Color Internegative Film are exposed through the appropriate separations using filter packs of the proper type.
Processing of the Color Negative, Color Print and Color Internegative Films is carried out in conventional type continuous processing machines, which provide for all of the steps required. These include, in addition to the washing steps, prebath for backing removal, color development, first fixing bath bleach, second fixing bath, and wetting agent or stabilizing bath. Processing of the Color Internegative Film requires the same solutions as for the Color Negative, but a somewhat shorter development time is used for the former. For the Color Print Film, a different color developer solution is used than for the Color Negative Film. Other solutions are the same.
The accompanying schematic diagram illustrates the various steps in processing Eastman color films with the Eastman Color processing machine. Two additional diagrams show details of the power buffer, which removes the anti-halation backing on the films, and the sound track developing unit.
Development of the sound track is done by means of a bead-type applicator wheel, which dips into a small tray of sound track developer solution. This unit is located in the production line of the processing machine, after the wash tank following the bleach tank.
The processed color internegative may be intercut with the original color negative for release printing onto Color Print Film, Type 5382. Reduction prints can also be prepared on the Type 7382 (16mm) Film from the 35mm color negative or internegative.
In order to maintain proper control of the processing of these color films, it is necessary to make regular chemical analyses of the various solutions and sensitometric tests as well.
More complete data concerning these new color films, including full details on processing procedure, formulae, etc., are contained in a paper to be published shortly in the Journal of the SMPTE, and prepared by W. T. Hanson, Jr., of the Research Laboratories, Eastman Kodak Company, Rochester. N. Y., and W. I. Kisner, Motion Picture Film Dept., Eastman Kodak Company, Rochester, to whom the author is indebted for the technical data furnished for this article. ”
(Foster, Frederick (1953): Eastman negative-positive color films for motion pictures. In: American Cinematographer, 34,7, July 1953, pp. 322-333, 348.)
The Eastman Color Process is a three-color subtractive color process introduced in 1950 by the Eastman Kodak Company.1 When it was introduced the process consisted of two elements that could be used singly or together: a coupler-incorporated, integral mask, three-color negative, and a coupler-incorporated, three-color print film. The negative achieved almost immediate success and has become the basic unit for most of the 35 mm color motion pictures photographed in the United States. The print film was used first chiefly for daily prints but its popularity has consistently increased and at present it is one of the most popular films used for motion picture release printing. A large part of the success of the Eastman Color Films has been due to the continuing effort of the Kodak Research Laboratories. Since the introduction of the process there have been six different camera negative films and seven different print films; also two different panchromatic separation films, four different internegatives, a reversal color intermediate which produces a duplicate negative directly from an original color negative, and a color intermediate film which can be used as either an intermediate positive or an intermediate negative.
During the first few years following its introduction Eastman Color Negative Film, Type 5247 was used to photograph feature films released in several print processes.
The Redskin Rode, Super Cinecolor
The Texas Rangers, Super Cinecolor
The Barefoot Mailman,Super, Cinecolor
Hurricane Island, Super Cinecolor
Sunny Side of the Street, Super Cinecolor
The Magic Carpet, Super Cinecolor
The Lion and the Horse, Warner Color
Carson City, Warner Color
The Miracle of Our Lady of Fatima, Warner Color
Springfield Rifle, Warner Color
House of Wax, Warner Color
Stop, You’re Killing Me, Warner Color
She’s Back on Broadway, Warner Color
The Sword of Monte Cristo, Super Cinecolor
The Robe, Technicolor
How to Marry a Millionaire, Technicolor
Beneath the Twelve Mile Reef, Technicolor
Drums in the Deep South, Super Cinecolor
Fair Wind to Java, Trucolor
Lady Wants Mink
The camera negative films manufactured under the name of Eastman Color Negative are multilayered color films which consist of three light-sensitive emulsions sensitized to red, green and blue light respectively, and coated on a single film support (Fig. 50). Incorporated in the emulsion layers are dye couplers which react simultaneously during development to produce a separate dye image in each layer complementary to the sensitivity of the layer. The light and dark areas of the image are reversed with respect to those of the original subject. Also the various color areas of the negative are complementary in color to the corresponding areas in the original scene. The overall orange cast of the negative film is due to the color contributed by the unused coupler in the red and green sensitive layers. These colored couplers act as masks to correct for the deficiencies of the dyes which form the image.
The dyes used in color photography, whether they are produced by dye coupling or by any other means, have absorption characteristics which lead to undesirable results when a color reversal film is duplicated or a color negative film is printed to a color positive film. In addition to the desired absorption, there is a certain amount of unwanted absorption in other regions of the spectrum which tends to degrade the color reproduction. The cyan-dye image absorbs some blue and green light in addition to the red light which it is intended to absorb, the magenta-dye image absorbs some blue light in addition to the green light which it is intended to absorb and the yellow-dye image absorbs some green light in addition to the blue light which it is intended to absorb. The effects of this unwanted absorption can be minimized through the use of separate silver masks as described by T. H. Milter.2 However, this type of correction is most suitable for still photography; when applied to motion picture photography it becomes both cumbersome and expensive. A more suitable method for motion picture reproduction is to use couplers which have an original color that is nearly equivalent to the unwanted absorption of the dye images found after coupling.3 The use of this technique to obtain improved color reproduction in motion pictures represents a significant advance in the state of the art. Couplers of this type were used for the first time in a motion picture film in the red and green sensitive layers of Eastman Color Negative Film. The cyan dye-forming coupler is colored orange. When it is converted by exposure and development to cyan dye the orange color is destroyed in proportion to the amount of development which takes place. The unexposed areas where no development takes place retain their original color forming an orange-colored positive image along with the negative cyan image in the red sensitive layer. The magenta dye-forming coupler is colored yellow. When it is converted by development to magenta dye the yellow color is destroyed in proportion to the amount of development which takes place. The unexposed areas where no development takes place retain their original color forming a yellow color positive image along with the negative magenta image in the green sensitive layer. The yellow dye-forming coupler is colorless because at the time when the process was developed there were no entirely suitable magenta colored couplers. If such a coupler were available it would be possible to improve the reproductions of yellow and green.
Six different color negative films having the general characteristics described above have been manufactured as Eastman Color Negative Films:
Eastman Color Negative Safety Film, Type 52474
A color negative film balanced for use under average daylight conditions, that is for the mixture of sunlight and some blue skylight. The speed of this film was E.I. 16, introduced in 1950.
Eastman Color Negative Film, Type 52485
A color negative film balanced for use with 3200°K Tungsten illumination. The speed of this film was E.I. 25 Tungsten and 16 Daylight with a Wratten 85 Filter. Introduced in 1953 this film replaced the earlier type 5247 film. In addition to having a higher speed for tungsten exposure it had finer grain and improved color reproduction compared to its predecessor.
Eastman Color Negative Film, Type 52506
A color negative film balanced for use with 3200°K Tungsten illumination. The speed of this film was E.I. 50 Tungsten and 32 Daylight with a Wratten 85 Filter. Introduced in 1959 this film replaced the earlier type 5248 film. Its improved characteristics included one camera stop more film speed and a new yellow-forming dye-coupler which produced an image dye having less absorption for green light, giving improved color reproduction.
Eastman Color Negative Film, Type 52517
A color negative film balanced for use with 3200°K Tungsten illumination. The speed of this film was E.I. 50 Tungsten and 32 Daylight with a Wratten 85 Filter. Introduced m 1962 this film replaced the earlier Type 5250 film. Its improved characteristics were substantially improved color balance.
Eastman Color Negative Film, Type 52548
A color negative film balanced for use with 3200°K Tungsten illumination. The speed of this film was E.I. 100 Tungsten and 64 Daylight with a Wratten 85 Filter. Introduced in 1968 this film replaced the earlier Type 5251 film. Its improved characteristics included one lens stop more film speed with no increase in graininess maintaining the same sharpness and preserving the same color rendition.
Eastman Color Negative Film, Type 5247/72479
A color negative film balanced for use with tungsten illumination. The speed of this film was E.I. 100 Tungsten and 64 Daylight with a Wratten 85 Filter. Introduced in 1972 this film replaced the earlier Type 5254 film. Its improved characteristics include a significant increase in sharpness, and reduction in graininess along with a substantial reduction in processing time.
The print films manufactured under the name of Eastman Color Print Film are multilayered color films which consist of three light-sensitive emulsions sensitized to blue, red and green light respectively and coated on a single film support (Fig. 51). Incorporated in the emulsion layers are dye couplers which react simultaneously during development to produce a separate dye image in each layer. When printed from a color negative or additively from three color-separation negatives a three-color reproduction of the original subject is obtained. The couplers used in these films are colorless.
Eastman Color Print Safety Film, Type 528110 (35 mm), 7381 (16 mm)
A color positive film balanced for printing with filtered tungsten illumination. The manufacturer’s instructions stated that better contrast and color reproduction could be obtained if this film was printed by additive rather than subtractive means. Introduced in 1950.
Eastman Color Print Film, Type 538211 (35 mm) and 7382 (16 mm)
A color positive film balanced for printing with filtered tungsten illumination. Introduced in 1953 this film replaced the earlier Type 5381 film. Its improved characteristics included better sharpness, better sound track and a new magenta dye-forming coupler which resulted in improved red reproduction.
Eastman Color Print Film, Type 738312 (16 mm)
A color positive film balanced for printing with filtered tungsten illumination. Introduced in 1959 this film was only available in the 16 mm size where its improved sharpness could be used with the greatest advantage.
Eastman Color Print Film, Type 538513 (35 mm) and 7385 (16 mm)
A color positive film balanced for printing with filtered tungsten illumination. Introduced in 1962 this film replaced the earlier Type 5382 (35 mm) and Type 7383 (16 mm). Its improved characteristics included better sharpness in the 35 mm size and higher effective speed and improved color balance in both 35 mm and 16 mm sizes.
Eastman Color Print Film, Type 738014
A color positive film balanced for printing with filtered tungsten illumination. Introduced in 1968 this film was only available in small format sizes where its improved sharpness and finer grain characteristics could be used with the greatest advantage. Since the improvements mentioned above were accompanied by a substantial loss in film speed a demand never developed for this film in larger formats.
Eastman Color Print Film, Type 538115 (35 mm) and 7381 (16 mm)
A color positive film balanced for printing with filtered tungsten illumination. When it was originally introduced in 1970 this film was only available in small format sizes (8 mm and Super 8). In January 1972 it became available in 35 mm formats. Its improved characteristics included sharpness and grain levels equal to Type 7380 with a speed increase equal to 5385/7385; these improvements were particularly significant in laboratories making high-speed reduction prints in 8 mm or Super 8 formats. A later improvement introduced in mid-1974 improved the yellow color reproduction.
Eastman Color Print Film, Type 5383 (35 mm) and 7383 (16 mm)16
A color positive film balanced for printing with filtered tungsten illumination. Introduced in 1974, this film was sold concurrently with Type 5381/7581. Its general characteristics were similar except this film had a hardened emulsion which permitted it to be processed in the higher temperature more rapid access ECP-2 process.
Approximately one year after the introduction of the first Eastman Color Negative and Print Films the system was made complete by the addition of two intermediate films:
1. A black and white panchromatic film having a contrast range intermediate between the negative and positive films usually used in black and white work.
Eastman Panchromatic Separation Film, Type 521617
The emulsion of this film was slower, sharper and had finer grain structure than the usual panchromatic duplicating films. Its purpose was for making three-color separation positives from color negatives. Introduced in 1951.
Eastman Panchromatic Separation Film, Type 523518
A black and white panchromatic duplicating positive film used for making three-color separation positives from color negatives. Introduced in 1956 this film replaced the earlier Type 5216 film.
2. A multilayered duplicating color negative film which consisted of three light-sensitive emulsions sensitized to red, green and blue light respectively (Fig. 52). Incorporated in the emulsion layers were dye-forming couplers which react simultaneously during development to produce separate dye images in each layer. This film differed significantly from its companion camera film Type 5247 in that the dyes formed were not complementary to the sensitivity of the emulsion layers. The magenta dye-forming coupler was in the top blue-sensitive layer so that the blue light exposure led to the formation of magenta dye. The cyan dye-forming coupler was in the middle green-sensitive layer so that the green light exposure led to the formation of cyan dye. The yellow dye-forming coupler was in the bottom red-sensitive layer so that the red light exposure leads to the formation of yellow dye. This arrangement of dye forming couplers and sensitizing was used to give maximum sharpness.
The internegative contained colored couplers similar to those used in Eastman Color Negative which provided automatic masking to correct in part for the unwanted absorption of the cyan and magenta dyes of the negative.
Eastman Color Internegative Safety Films, Type 524319
A multilayered duplicating color negative film which consisted of three light-sensitive emulsions sensitized to red, green and blue light respectively. Introduced in 1951.
Eastman Color Internegative Film, Type 524520
A multilayered duplicating color negative film which consisted of three light-sensitive emulsions sensitized to red, green and blue light respectively. Introduced in 1953 this film replaced the earlier Type 5243 film. Its general characteristics were similar to those of its predecessor with improvements in graininess and sharpness.
In 1956 the Eastman Color system was augmented by the addition of two new films, Type 5270 a duplicating negative for use with reversal color originals and Type 5253 a color intermediate that can be used as either a color master positive or a duplicate negative. The latter eventually replaced completely the Type 5245 internegative.
Eastman Color Internegative Film, Type 527021 (35 mm) and 7270 (16 mm)
A slow, fine-grain, long latitude multilayered duplicating color negative film which consisted of three light-sensitive emulsions sensitized to red, green and blue light respectively (Fig. 53). As in the Eastman Color Negative Film colored dye-forming couplers were employed to provide color correction for unwanted absorption of the image dyes of the reversal original.
Eastman Color Internegative Film, Type 527122 (35 mm) and 7271 (16 mm)
A slow, fine-grain, long latitude multilayered duplicating color negative film for use with reversal camera originals. Introduced in 1968, this film replaced the earlier Type 5270/ 7270 film. Its improved characteristics included improved sharpness, finer grain, improved color reproduction, slightly higher speed, improved curve shape and increased processing compatibility. Optimum processing is obtained in the Eastman Color Print Process at 75°F using the same solution composition, times, and temperatures required to process Eastman Color Print Film.
Eastman Color Intermediate Film, Type 525323 (35 mm) and 7253 (16 mm)
A multilayered duplicating color film that can be used for the preparation of either a color master positive or a color duplicate negative. In addition to the color-forming coupler this film contains absorbing dyes which prevent scattered light traveling within the emulsion layers. This improves image sharpness to the degree that it is not necessary to use the inverted order of layers employed in Types 5243 and 5245 (Fig. 54).
As advances in emulsion technology continued to progress it became possible to design a new film which by reversal processing could produce a duplicate negative in a single printing and processing operation. This advancement improved laboratory efficiency but more important to the final screen image it also improved graininess, color reproduction and sharpness. Introduced in 1968 the new film was an immediate success.
Eastman Color Reversal Intermediate Film, Type 524924 (35 mm) and 7249 (16 mm)
A multilayered incorporated-coupler duplicating color film which by reversal processing produces duplicate color negatives directly from the original negative. The spectral sensitivity and image dye absorption characteristics are similar to those of Eastman Color Intermediate Type 5253 permitting intercutting with the original and other color negative duplicating films. With this new film it is necessary to place the original base to emulsion in an optical printer when exposing the duplicate. This is required to preserve orientation as well as to introduce effects or change picture format, also modified techniques are required for fades and dissolves. An additional requirement for the use of this new film is the installation of a new processing machine and process CRI-1.
Eastman Color Processing
The films employed in the Eastman Color family can be processed in conventional type continuous processing machines with minor modifications to allow for all of the steps required. Until the introduction of the ECN-2 process in 1972 the processing steps had remained approximately the same throughout the evolution of the process. Improvements as they came were accomplished in film manufacturing. The only significant exception was the change in processing times necessary when the film was changed from Type 5247 to 5248. The introduction of EC N-2 process in 1972 had a major effect on motion picture laboratory operation. The processing time was reduced from approximately 45 minutes to 12 minutes, also substantial changes were made in solution chemistry. The result was more rapid access to the developed negative and a reduction of processing machine size for a given level of productivity. A more gradual evolution has taken place in the print process. Changes in film manufacture made possible increases in solution temperature with subsequent reduction in processing time.
The following tables give the processing steps and times for the various members of the
Eastman Color Process family described in the previous sections.25 Table 1 shows the ECN- 1 process for color negatives and internegatives, the reduced times for the higher temperature ECN-2 process being given in Table 2. The steps of the positive process ECP-1 are shown in Table 3, with the EC P-2 process in Table 4. Table 5 indicates the steps of the reversal process CRI-1.
1 HANSON, W. T., “Color Negative and Color Positive Film for Motion Picture Use,” Journal of The Society of Motion Picture and Television Engineers, March 1952, p. 223,
2 MILLER, T, H., “Masking: A Technique for Improving The Quality of Color Reproductions.” Journal of The Society of Motion Picture Engineers, Feb. 1949, pp. 133-155.
3 HANSON, W. T., “Color Correction With Colored
Couplers,” Journal of The Optical Society of America 40, March 1950, pp. 166-171.
4 HANSON, W. T., “Color Negative and Color Positive Film for Motion Picture Use,” loc. cit. pp.
5 HANSON, W. T. and KISNER, W. I., “Improved Color Films for Color Motion Picture Production,” Journal of The Society of Motion Picture and Television Engineers, Dec. 1953, pp. 670-680.
6 DUNDON, M. L. and ZWICK, D. M., “A High Speed Color Negative Film,” Journal of The Society of Motion Picture and Television Engineers, Nov. 1959, pp. 735-738.
7 KISNER, W. I., “A New Color Negative Film for Better Picture Quality,” Journal of The Society of Motion Picture and Television Engineers, Oct. 1962, pp. 776-779.
8 BEELER, R. L., MORRIS, R. A. and WESTON SIMONDS, C, “A New Higher Speed Color Negative Film,” Journal of The Society of Motion Picture and Television Engineers, Sept. 1968, pp. 988- 990.
9 ANDERSON, R. G., BONHEYO, G. L., CLIFFORD, J. D., DANIELSON, A. D., HESTER, J. R. and SHAFER, R. K., “Improved Emulsion and Processing Technology for Motion Picture Negative Film.”
10 HANSON, W. T., “Color Negative and Color Positive Film for Motion Picture Use,” loc. cit., pp. 231-238.
11 HANSON, W. T. and KISNER, W. I., “Improved Color Films for Color Motion Picture Production,” loc. cit., pp. 681-692.
12 KISNER, W. I., “A Higher Speed Color Print Film,” Journal of The Society of Motion Picture and Television Engineers, Oct. 1962, p. 779.
13 Ibid., pp. 779-781.
14 “New Products,” Journal of The Society of Motion Picture and Television Engineers.
15 “New Products,” Journal of The Society of Motion Picture and Television Engineers.
16 SCHAFER, R. K., BAPTISTA, J., O’CONNELL, R. and KNUTSEN, E. V., “A New Color Print Film with a Shortened Processing.” Presented at 115th Technical Conference of The Society of Motion Picture and Television Engineers, Los Angeles, April 1974.
17 ANDERSON, C, GROET, N. H., HORTON, C. H. and ZWICK, D., “On Intermediate Positive Internegative System For Color Motion Picture Photography,” Journal of The Society of Motion Picture and Television Engineers, March 1953, pp. 217-225.
18 Internal Memo, Motion Picture Film Department – Eastman Kodak Company, 1956.
19 ANDERSON, C. et al., “An Intermediate Positive Internegative System for Color Motion Picture Photography,” loc. cit., pp. 217-225.
20 HANSON, W. T. and KISNER, W. I., “Improved Color Films for Color Motion Picture Production,” loc. cit., pp. 694-696.
21 ZWICK, D. M., BELLO, H. I. and OSBORNE, C. E., “A New Color Internegative Film for Use in Color Motion Picture Photography,” Journal of The Society of Motion Picture and Television Engineers,
August 1956, pp. 426-428.
22 BROWN, R. C, MORRIS, R. A. and O’CONNELL, R. J., “An Improved Color Internegative Film,” Journal of The Society of Motion Picture and Television Engineers, Sept. 1968, pp. 990-994.
23 BELLO, H. J., GROET, N. H., HANSON, W, OSBORNE, C. E. and ZWICK, D. M., “A New Color Intermediate Positive Intermediate Film System for Color Motion Picture Photography,” Journal of The Society of Motion Picture and Television Engineers, April 1957, pp. 205-209.
24 BECKETT, C, MORRIS, R. A., SCHAFER, R. K. and SEEMANN, J. M., “Preparation of Duplicate Negatives Using Eastman Color Reversal Intermediate Film,” Journal of The Society of Motion Picture and Television Engineers, October 1968, pp. 1053-1056.
25 “Production of Motion Pictures in Color Using Eastman Color Films.” (Motion Picture Film Department, Eastman Kodak Co., Rochester, N.Y., 1952, 1954, 1960, 1963.)”
(Ryan, Roderick T. (1977): A History of Motion Picture Color Technology. London: Focal Press, on pp. 148-158.)
“The Evolution of Eastman Color Motion Picture Films*
By Wesley T. Hanson, Jr.
The author became active in color film design shortly after the introduction (1935) of the Kodachrome amateur reversal film, created by Mannes and Godowsky. In this article, he traces the history of Eastman Color films from Troland’s patent on monopack film and Fischer’s discovery of color couplers, to the completion of the first negative-positive Kodacolor film system for still photography.
The author of this paper was involved in the research and development work related to motion picture products for many years, and feels very strongly that he is still a member of the motion picture community. Having been closely associated with much of the work that led to the Eastman color motion picture film, I shall relate some of my recollections of the highlights of that work.
However, this will not be a complete history of the development of color motion pictures. Little reference will be made to work of other organizations active in this field. In no way is this omission intended to belittle the importance of those other activities.
Kodachrome – 1935
In 1935, Kodachrome 16-mm amateur film1 was placed on the market. This was the first commercially available integral subtractive color process, and it was immediately successful. I joined the program in 1936 in a group whose responsibility it was to apply the process to 35-mm slides. The process was ideal for this purpose, giving excellent transparencies of very low grain and high sharpness. From the beginning, one of our objectives was to adapt this process for professional motion picture purposes.
But there the requirements were, and are, quite different. Many prints, or copies, are required, and intermediate steps in going from the camera film to the print are required for effects and embellishments. Copying the dye image of Kodachrome, or any other color original, introduces problems, the details of which we shall not consider at this point. Masking techniques were the then prevalent approach to solving those problems, and we investigated every way we could think of to mask a Kodachrome film.
Monopack Color Film
At that time, the Technicolor process with its three-strip camera was the only commercial motion picture color process. In 1931, Troland2 of Technicolor had been granted a patent on “monopack” films, a single film base on which were coated the many layers of emulsion required in a color film. Kodachrome was such a film.
Because the Technicolor process involved three separation negatives exposed in a camera, it was felt that a monopack film using the Kodachrome process could be used by Technicolor with separation negatives being made from the original film. Research on evolving a film which would have the characteristics suitable for such use was undertaken and monopack was worked out. In 1942, Twentieth Century Fox released the picture Thunderhead, shot on monopack and printed in Technicolor.3 Monopack was used for a few years, and in the late 1940’s Cinecolor – who by then had developed a three-color process – and some other companies were giving consideration to shooting some motion pictures in monopack.
At the same time that the work on monopack was being done, general research on color photography was also progressing. lt became possible to incorporate color couplers directly into the emulsions and coat them in the several superimposed layers instead of having the couplers in the developers, as was the case with Kodachrome. This would lead to a much simpler process.
The idea was first put forth by Fischer in Germany in 1912, but it was not until the late 1930’s that it became possible to carry it out. Agfa in Germany worked out incorporated couplers of one particular type. In the Kodak Labs, couplers which were dissolved in an oily liquid and then dispersed in gelatin, were developed. This was called the protected coupler process.4
A Still-Photography Negative-Positive Color Process
The incorporation of protected couplers made possible a negative-positive color process for still photography and, in 1942, Kodacolor roll film5 for amateur use was placed on the market. Prints made from these color negatives had color problems of their own, resulting from the characteristics of all the intervening dyes, just as copies of Kodachrome slide had theirs. A great deal of work aimed at solving these problems was carried on. In 1943, colored couplers6 were invented, and these essentially solved the problem. Many papers have been presented describing how colored couplers effect color correction by appropriate masking, and hence we need not consider this aspect here. When such couplers were finally perfected for use in negative sheet and roll films, the quality of the prints was very much improved.
lncorporated couplers were also soon used in a reversal material and process, the Ektachrome sheet film. Thus, in the 1940’s Kodacolor roll film and Ektachrome sheet film were available for still photography.
Motion Picture Color Film – A Dream
At that time, the application of incorporated couplers to professional motion picture film was not feasible because of speed, sharpness, and graininess problems. Coatings containing couplers were very thick. But it did appear that incorporated couplers in a very fine-grain print film emulsion might give adequate grain and sharpness, though at very low speed, for use by Technicolor in making rush prints from their camera negatives. Work on this kind of print film was initiated.
But we still had dreams of an entire Integrated motion picture process. In considering such a process we had to keep three major factors in mind: quality, economics, and convenience. There had to be at least; a camera film, a print film, and an intermediate stage, and we dreamed of the day when films with incorporated couplers would have adequate graininess and sharpness.
So we speculated. If we used an Ektachrome camera film, we could have an internegative and a positive; or we could use three stages of reversal intermediate and a positive film…and so on. We were probing: what was the best combination of the technologies available?
One interesting process that we were working on was “monolayer”. It had become possible to separately sensitize three emulsions to red, green, and blue, and to mix them and coat them in one layer without the sensitizers wandering.7 The processing of such a film fairly complicated, but the economics of the film itself would be the most desirable of any we could imagine. This might be a low-cost release print film. For release print film, with hundreds of prints being made, film cost was of prime importance.
Such a film, using a reversal process, would require three stages of reversal if Ektachrome were the camera film. If Kodacolor were the camera film, then an interpositive would be the intermediate stage. Exploratory experiments in all these directions were carried out during the late 1940’s.
By 1949 colored couplers had been introduced into Kodacolor roll film. The print quality was excellent. Enough progress had been made in improving the grain characteristics and sharpness, and therefore we decided to take a look at it as a possible motion picture film. Some of the products coated for roll film experiments was prepared in the 35-mm format Negatives exposed on this material were printed on experimental positive film which was being worked on for other purposes.
When we first viewed the results we were very excited. We immediately showed the prints to company management. the color and tone qualities were better than anything else we had seen in a two-stage print process. They were as good or better than those of Kodachrome original. The graininess and sharpness could stand improvement, but still we were very encouraged. We decided to institute a major development program to work out a color negative camera film and a color print film. Many problems had to be solved and the question of intermediate was not even addressed at the time. We would solve that in due course.
Improvement of Graininess
The graininess and sharpness problems would not go away. Finally an idea occurred to us. Graininess would be improved if each of the three color components of the negative was separated and coated in two layers, each layer being much thinner than previous coatings. The first tests of this idea, though all out of color balance and showing many defects, did show, however, that we were on the right track with regard to graininess.
To obtain adequate sharpness and finer grain in the print film, it was decided to put the magenta layer on top because the magenta dye image carries a major part of the sharpness. Yellow, contributing least to the sharpness, would be coated on the bottom. Chlorobromide emulsions which could be used in this format were worked out and a print film with very high resolving power was produced.
Motion pictures projected in our small viewing room and viewed from quite close up looked excellent. However, when they were presented to a large audience on a big screen and viewed at greater distance, where the magnification was actually less than it was on our small viewing screen. The pictures were very unsharp. Back to the drawing board!
What was the problem? It turned out to be back scatter of light from the bottom blue sensitive layer. This did not affect resolving power but resulted in very low acutance at low line frequencies. We added inert absorbing dyes to the coating and eliminated this problem.
Eastman color negative and print films were described in a paper presented to the SMPTE in 1950 and published in the JOURNAL of the SMPTE in I952.8
In 1950, Columbia Pictures photographed The Redskin Road on Eastman color negative film. It was released by Cinecolor in 1951.
The first full-length color motion picture using the entire Eastman color process, negative and print films, was Royal Journey, which was produced by the National Film Board of Canada. This feature presented the royal tour made by Her Majesty Queen Elizabeth, who was then Princess Elizabeth, and the Duke of Edinburgh. Twenty-eight days after the end of the tour, production was complete and a test print was ready. The film was released by Columbia Pictures of Canada 12 days later on December 21, 1951.
State of the Art in the 1950’s
The color negative film at that time had an Exposure Index of 16 and was balanced for use with daylight type arc lights. All color motion pictures were shot with this type of illumination at that time. The paper describing the negative and the print films mentioned the possible use of separation positives and separation negatives, but did not deal seriously with the major requirements of intermediate steps.
An integrated process involving a negative, an interpositive and internegative, and a positive was still needed. Short of this, the requirements for motion picture production could be met by means of a process involving a negative, separation positives, an internegative, and a positive. The optical effects and other embellishments, printed by means of separation positives on a single internegative, could lead to an internegative which could be intercut with the camera negative so that release printing could be done in a continuous and economical fashion. Because a simple internegative film which could be printed from black-and-white separation positives had much less stringent design requirements than a color film which could be used as both an interpositive and an internegative, design of the simpler internegative film became the objective
As our research continued we concentrated on emulsions, on tone reproduction, on latitude, on graininess, on sharpness. Some of this research was aimed at high-speed emulsions for negatives. Some was aimed at finer grain emulsions for use in intermediates or print film.
For an internegative film, which was to be printed from separation positives, the magenta layer could be on top. This would give the best sharpness. The green separation positive could be printed with blue light on a top, blue sensitized layer that gave a magenta dye image. So, such a “false sensitized” intermediate was conceived. Work on this, as well as work on improving the negative was successful and, in 1953, an entire integrated process was presented. The color negative, separation positives, false sensitized internegative, and release print film were described in an SMPTE paper.9
The camera film had an Exposure Index of 25 and was now balanced for use with tungsten illumination. This was the first time that professional color motion picture film had been fast enough to be used with tungsten lights. Here was a complete color system to be used widely by the motion picture industry.
Simultaneous Multilayer Coating
There was still much left to be desired, however. We still had our dreams. Monolayer Kodachrome-type film continued to intrigue us as a very economical release print film. A process involving a negative, and interpositive, and a monolayer film still demanded some attention. Then another invention came along which changed all of that.10 In early 1954, the idea evolved that the various layers in a color film could be coated simultaneously – without mixing – by the proper arrangement of the various emulsion coating outlets delivering the emulsions to the film base during the coating operation. Believe it or not, this idea worked! Several emulsions could be coated together without intermixing as they flowed on the film base. This changed things. The economics of a multilayer print film could not compare much more favorably than before with a monolayer film. Because of the monolayer processing problems, and because of the inadequate quality so far achieved, work on monolayer was discontinued.
Research would now be concentrated on developing an interpositive film which would replace the separation positives. This work was successful. In a paper presented at the SMPTE meeting in 1956 and published in 1957,11 a new intermediate film was described having a gamma of one and sufficient latitude for making high-quality intermediates. It could be used as an interpositive and printed back on itself to make a duplicate negative.
Perfectioning Motion Picture Color Films
Now there was an integrated color film system that would fill the requirements of the motion picture industry. The production of motion pictures in color did not have to be restricted any longer to a few experienced organizations but could be undertaken wherever desired.
However, much still remained to be done and the research effort was not relaxed. In 1959, a paper titled “A High-Speed Color Negative Film” was presented.11 This film had an Exposure Index of 50.
One of the ongoing prime objectives of research on photographic emulsions was, and is, to improve the speed-to-grain ratio. Higher speed emulsions can be made by using larger silver halide grains, but the graininess becomes unacceptable. So the basic improvements involve improvements in the ratio of speed to grain. As a result of continuing basic research in emulsion technology, the graininess of the 50-E.I. Eastman color negative film was improved in 1962.13 At the time this product was announced, a new print film was also announced,13 which had a fourfold speed increase and also had improved graininess and color. These qualities resulted from many years of research in emulsion technology, couplers, sensitizers, processing chemistry, and coating technology.
Nineteen sixty-eight was a banner year. A new, higher speed color negative film with an Exposure Index of 100 was announced.15 That same year, a new kind of duplicating film, “color reversal intermediate,” was made available.16 Duplicate color negatives are shipped to various countries for release printing in those countries. A color reversal intermediate film is the means for making, in one single step, a duplicate color negative for such a purpose from an original camera negative.
The idea for this film came from the Kodak Laboratories in France, and the original work on it was done there. In this film, colored couplers are used for the purpose of giving improved color reproduction. This is the only reversal film using such couplers. The emulsion was designed for processing in the Ektachrome reversal process which was already in use in the 16-mm commercial field.
It should be noted that the development of the use of color in the 16-mm field is an entirely separate story with which this report docs not deal. However, 1968 did bring the two fields together, with the introduction of 16-mm internegative film.17 This product was designed for printing from reversal color originals on Eastman color print film.
The characteristics of this film were influenced by the use of a new kind of coupler, a development inhibitor releasing or DIR coupler. Such color forming couplers were invented back in the early 1960’s18 and it took a long time to tame them, to get them under control for use in a photographic product. Such couplers cause edge effects and enhance image sharpness. This is quite important in 16-mm film.
Research continued on emulsions, on couplers, on hardeners, and on many other aspects of photographic systems. All of this research came together in 1972. At the 112th Conference of the SMPTE a new color negative process, Eastman Color Negative Process 2, was described, along with the new Eastman Color Negative II film.19 This film had the same speed, 100 E.I., as the previous film but had improvements including better grain quality. These improvements made it useful in the 16-mm field.
Research on the color positive film took a different turn. For this film, economics is of prime importance. Productivity gains are important. The work on new hardeners led to the ability to work out a film which could be processed at higher temperatures and thus in shorter times. In 1973, Eastman color print film for a short process was introduced. This process takes nine-and-a-half minutes for completion, and it is the author’s understanding that some of the processing machines in use run at 400 feet per minute.
In 1976, the color intermediate film was improved and adapted for processing in the ECN2 process. That just about brings us up to date. Now, most motion pictures are photographed in color. This procedure is used worldwide. Duplicate negatives are shipped all over the world to be printed locally.
Some Final Words
We have come a long way from the 16-E.I. daylight film with marginal graininess and sharpness to a 100-E.I. tungsten balanced film for general purpose use, with quality intermediates and print film adapted for high speed printing and processing. It is not the author’s intention to give the reader the false impression that each of the described breakthroughs came easy. On the contrary! Each research project, each invention and innovation in film design represents an enormous commitment. Making each invention work, solving the many problems of tone reproduction, color balance, stability, which are introduced when each of the many variables is changed, has required blood, sweat, tears, and faith. And it won’t stop here.
What is to come? I shall not try to predict improvements in detail, but I feel quite confident that many improvements will be made so that the utility and convenience and the relative economics of motion pictures in color will continue to improve. Motion pictures and the motion picture industry will continue to grow in importance and will have a pervasive influence and an enormous impact on our culture and our social behavior. As the world becomes more and more visually oriented, motion pictures will influence the way we perceive it through news, through social commentary, and through entertainment.
*A paper presented at the 122nd annual SMPTE Technical Conference, November 9-14, 1980, New York, N.Y. Author: Vice President (Ret.) and Director, Research Laboratories, Eastman Kodak Company, Rochester, N.Y.
1 Mannes L. D., Godowsky L. Jr., “The Kodachrome Process for Amateur Cinematography in Natural Colors,” J. SMPE, 25: 65-68, July 1935.
2 Troland L. T., U.S. Patent 1,808,583, 1931.
3 Clarke C. G., “Practical Utilization of Monopack Film,” J. SMPE, 45: 327-332, Nov. 1945.
4 Jelley E., Vittum P. W., U.S. Patent 2,322,027, 1943.
5 Mees C. E. K., “Direct Processes for Making Photographic Prints in Color,” J. Franklin Inst., 233: 41-50, Jan.1942. (Also in J. SMPE, 42: 230-238, Apr. 1944.)
6 Hanson W. T. Jr., U.S. Patent 2,449,966, 1948.
7 Carroll B. H., Hanson W. T. Jr., U.S. Patent 2,592,243, 1952.
8 Hanson W. T. Jr., “Color Negative and Color Positive Film for Motion Picture Use,” J. SMPE, 58: 223-228, Mar. 1952.
9 Hanson W. T. Jr., Kisner W. I., “Improved Color Films for Color Motion Picture Production,” J. SMPE, 61: 667-701, Dec. 1953.
10 Russell T. A., U.S. Patent 2,761,791, 1956.
11 Bello H. J. Jr., Groet N. H., Hanson W. T. Jr., Osborne C. E., Zwick D. N., “A New Color Intermediate Positive – Intermediate Negative Film System for Color Motion-Picture Photography,” J. SMPE, 66: 205-209, Apr. 1957.
12 Dundon M. L., Zwick D. M., “A High-Speed Color Negative Film,” J. SMPE, 68: 735-738, Oct. 1962.
13 Kisner W. I., “A New Color Negative Film for Better Picture Quality,” J. SMPE, 71: 776-779, Oct. 1962.
14 Kisner W. I., “A Higher Speed Color Print Film,” J. SMPE, 71: 779-781, Oct. 1962.
15 Beeler R. L., Morris R. A., Simonds C. W., “A New Higher Speed Color Negative Film,” J. SMPTE, 77: 988-990, Sept. 1968.
16 Beckett C., Morris R. A., Schafer R. K., Seemann J. M., “Preparation of Dublicate Negatives Using Eastman Color Reversal Intermediate Film,” J. SMPE, 77: 1053-1056, Oct. 1968.
17 Brown R. C., Morris R. A., O’ Connell R. J., “An Improved Color Internegative Film,” J. SMPE, 77: 990-994, Sept. 1968.
18 Barr Ch. R., Williams J., Whitmore K. E., U.S. Patent 3,227,554, 1966.
19 Barr Anderson R. G., Bonheyo G. L., Clifford J. D., Danielson A. D., Hester R. J., Schafer R. K., “Improved Emulsion and Processing Technology for Motion-Picture Color negative Film,” a paper presented at the 112th SMPTE Technical Conference, Oct. 22-27, 1972, in Los Angeles, Calif.”
(Hanson, Wesley T. (1981): The Evolution of Eastman Color Motion Pictures. In: Journal of the Society of Motion Picture and Television Engineers, Vol. 90, No. 9, 1981, pp. 791-794.)
No dye produced by colour coupling has perfect spectral transmission, showing unwanted absorbtion in certain regions; this can lead to a degree of colour degradation in absorbtions can be partly compensated by the use of masking images, which were originally separate low-density black and white prints, but Eastman Kodak developed a method by which the colour couplers in the emulsions could themselves form the required masks. For example, in the green sensitive layer, the magenta-forming coupler has a yellow colour; where a magenta image is formed, this yellow is destroyed but remains in the other areas to provide a proportionate correction for the unwanted absorbtion of blue light inherent in the imperfect magenta dye. Similarly in the red-sensitive layer, an orange-coloured coupler corrects for the unwanted absorbtion of blue and green light by the cyan image. Eastmancolor negative with this integral masking thus has a characteristic orange-yellow appearance in the unexposed areas, very different from the unmasked Agfacolor negative and its derivatives.
Eastmancolor negative and positive were first introduced in 1950, the print stock having the visually important magenta image as the top layer, unlike Agfacolor positive. But it was the appearance in 1953 of an improved negative balanced for exposure to tungsten light which rapidly led to its widespread adoption: within a year or so it had clearly become the most favoured material for professional cinematography. On the positive side, Eastman’s lead was by no means so well defined: Technicolor continued to make large quantities of release prints by dye-transfer, while Agfa-Gevaert, Ferrania (later 3M) and Fuji all produced competitive colour print stocks once the market was established.
In due course, the other manufacturers recognised the need for a masked negative stock but it was also clear that the laboratories could only conveniently process materials compatible with the developing systems installed for Eastmancolor. This was re-emphasised in 1974 when Eastman introduced revised processes for both negative and positive in which higher solution temperatures greatly reduced processing times; compatible stocks from both Agfa-Gevaert and Fuji followed.
As the table shows, there has been a steady increase in the sensitivity of colour negative, with lower lighting levels and costs, and this has been accompanied by continued improvement in image quality – finer grain, higher resolution, better tonal and colour reproduction. As a result, the need for large-area negative systems for wide-screen presentation has practically disappeared, while 16mm negative has widely replaced reversal stocks in many applications.
Similar improvements have taken place in the positive print materials, one of the most recent being better colour stability in long term storage, a factor which was sadly lacking in many of the earlier tripack stocks.”
(Happé, Bernard (1984): 80 Years of Colour Cinematography. London: British Kinematograph Sound & Television Society, on p. 16.)
“Blazing Technicolor”, “Stunning Trucolor”, and “Shocking Eastmancolor”
David L. Parker
Like almost everything else, the technical evolution of the movies did not happen in a simple, orderly way. It took trial-and-error experimentation to arrive at an industry standard for sound. Tussling over a standardized aspect ratio continues to this day. And the means of filling a film image with color have been many and diverse. Film preservation offers an opportunity to trace not only the aesthetic and sociological milestones (and detours), but also the technical, physical changes that film has gone through – a history both complicated and, more to the point, colorful as well.
Although it is commonly believed that films in color began with the introduction of full-color Technicolor in the middle thirties, from the outset there have been three kinds of color films, and examples of these may be found in The American Film Institute Collection and other collections in the Library of Congress.
As early as 1895, New York City audiences saw movies that had been colored by hand. The Edison Kinetoscope Company in their film of Annabelle’s Dance attempted to simulate the effects of the colored lights which played over her body during her stage performance by hand-coloring the black-and-white pictures frame by frame. A print of this 1894 film in the collection of the Library of Congress has her white dress hand-painted in half a dozen different colors.
By 1907, Pathé Frères was applying color mechanically, by the use of stencils, one cut for each color, then retouching by hand. The colors, though sometimes charming, were limited in number and in realism, but, thanks to them, a Pathe film such as Down in the Deep, could offer pink fairies rising from blue sea froth, green whales, and heroes in gold armor, an effect so striking that it was remembered from his childhood movie-going by the Russian director Sergei Eisenstein.
The practice of hand-coloring a part of the black-and-white picture persisted as late as 1922, as can be seen in the red-painted heart in a print of Cecil B. De Mille’s melodrama, Fool’s Paradise. Because the rest of the frame is without color, the colored part seems to jump out with three-dimensional ferocity and crudity. While no modern theatrical hand-tinting has been done, contemporary filmmakers such as Jules
Dassin and Claude Lelouch have effectively adapted other antique effects such as occasionally tinting and toning sections of their films in much the same way as was common in the first two decades of the century.
Tinting (the immersion in a dye of the developed black-and-white film by which the entire stock is colored uniformly) can be seen in the amber-colored love scenes and blue-colored night driving scenes of Lelouch’s A Man and A Woman (1966), while toning (chemical treating of the developed black-and-white image in which only the image is colored by altering the color of the silver deposit) survives in the brown-and-white images of the sepia night sequences of Dassin’s 10:30 PM Summer (1966). Such current uses of the technique may impress modern audiences unaware that feature film prints in the collection dating from the twenties are likely to be on film stock tinted any of eleven standard colors. This process was used functionally – blue for night scenes – and often dramatically, as in the sudden plunging of the slightly tinted lavender into deep red in the Nazimova Salomé of 1922.
When the emulsion coating on the film was tinted one color and its clear celluloid base was toned another, a two-color effect was achieved, as in the combination of amber and green in the Goldwyn romance, A Tale of Two Worlds (1922). The travel films and landscape shorts of the twenties attempted to use the tinting and toning effects to advantage as well, making the earth brown and the sky and water green.
In contrast to hand-coloring, and tinting and toning, “natural color cinematography” was attempted in 1909 with a process known as Kinemacolor. “Natural color” referred to the fact that Kinemacolor recorded directly on film from nature instead of applying colors artificially later. Its natural colors were virtually restricted to reds and browns, produced by the filming and projecting of its subjects through alternating filters while the frames on the film itself alternated between red and green. This effect presented to the eye, as the patent application worded it, “a satisfactory rendering of the subject in natural colors.”
Kinemacolor – spotty, smeary, and suffering from a rash – was a successful vaudeville “act” reaching its greatest commercial standing with the longest film produced before 1911, the two and one-half hour newsreel of the crowning of George V as Emperor of India at the Delhi Durbar, its ceremonies and pageantry recorded by twenty-three cameramen.
An improved Kinemacolor known as Prizma appeared in 1916. One side of its film was coated red-orange, the other coated blue-green to give a compromise coloring to its subject matter. Prizma’s OLD GLORY (1916) offers an accurate rendering of the United States flag, but gray skies and brown-gray grounds and buildings. A copy of Prizma’s A Day With John Burroughs (1917) captures insects, flowers, children, and the eighty-year-old naturalist in primary reds and greens.
That same year the Technicolor Corporation produced The Gulf Between, a feature-length fiction film in Jacksonville, Florida. Its unsuspenseful and drawn-out scenario was designed solely to show off the colors best reproduced by the process. This was achieved by placing fileters between the film and the screen during projection.
Although the principles underlying color photography were known as early as 1862, it would not be until the 1920s that a partially satisfactory color motion picture would be achieved. This came about as Technicolor dropped its additive color process (in which one color is added to another to produce a third color) and developed a subtractive process (in which colors are produced by filtering out other colors, each positive image being printed in a color complementary to the color of the filter). Technicolor’s two-color process was used for Metro’s Toll of the Sea, a two-reel variation on “Madame Butterfly” with Anna May Wong, which was expanded into a full-length, five-reel feature and released in 1923. The film was praised for its bright, yet non-harsh colors and its “nearer simulation of nature.” The film was a success, grossing $250,000.
The next year William van Doren Kelly, the developer of the Prizma process, introduced his subtractive process called Kellycolor. That Kellycolor was not as good as the perfected Prizma is shown by a 1927 two-reel drama in the process called The Gaucho.
Following the two-color Technicolor sequences in The Ten Commandments (1923) and Ben Hur (1925), it became the fashion to include color sequences in black-and-white films. Typically, color was not used to intensify dramatic development but to provide an interlude, as exemplified by the 1926 fashion show sequence of Irene; by the Corpus Christi Day processional at St. Stephen’s in Vienna with its plumes, epaulettes, and solemn cavalry in Stroheim’s The Wedding March (1928); and in thirty other films in a five-year period (1925-1930). Dramatically significant use of color would not appear until the last years of two-color Technicolor (1932-1933).
Mack Sennett, the pioneer comedy producer, introduced color in a subtractive red-and-green process he called Sennett Color into such 1928 two-reel comedies as The Swim Princess and one that survives, The Campus Vamp, featuring Carole Lombard. Within the next year, over forty shorts were produced in two-color processes in America. This tremendous increase in volume resulted in quality control problems as well as the more basic problems inherent in any two-color process: a two-color process yields a compromise coloring with less pure values than are possible with three-color systems, which can theoretically produce any color in the spectrum. Two-color processes could reproduce pinks, salmons, browns, tans, red-orange, and blue-green but could not reproduce pure reds and yellows, neutral grays, light or deep blues, violets and purples.
Tony Gaudio, cinematographer of the first all-sound, two-color musical On With The Show (1929), remembered that certain colors were achieved by showing the camera an entirely different color, that “an unbelievable amount of light was needed,” and that makeup was very unnatural.
Critics scoffed at “aluminum skies” and “revolutionary reds.” Color became identified with the decoration of drama and formula musicals, as typified by color sequences from The Dance Of Life (1929) – a Ziegfeld-like reel of production numbers – and from the remake of Mae Murray’s Peacock Alley (1930).
Perhaps the best surviving example of the two-color Technicolor process is The Mistery of the Wax Museum (1933), the Warner Brothers film recently restored by the AFI in collaboration with The Museum of Modern Art. It is still remarkable for the sharpness of its photography, its virtual absence of grain, and the rendition of its colors, but most especially for its use of a limited palette for effect: the lack of accuracy in its flesh tones was used to accentuate an eerie resemblance between the human actors and the museum’s wax models, crucial to the effectiveness of the plot. In such sequences, dramatic mood was created by colored light and two-color renditions of the subject matter, so that no lack in the coloring was felt by audiences. But it came too late to restore faith in two-color films.
The success of Disney’s three-color Technicolor cartoons, Flowers And Trees (1932), and The Three Little Pigs (1933), not only prompted Columbia to produce its “Krazy Kat,” “Scrappy,” and “Barney Google” cartoons entirely in Technicolor by 1935, but also launched the first successful full-color system. From this time three-color Technicolor would dominate until the introduction of Eastmancolor.
The film used to record the color information in three-color Technicolor was panchromatic silver bromide stock, each of the three black-and-white strips registering only the relative intensities of its primary color. In the special Technicolor camera, three black-and-white negatives, exposed at the same time behind their respective filters, combined to give a complete color record: a red barn in a green field photographed against a blue sky would leave on the red record negative an image of the barn; on the green record black-and-white negative an image of the field; and on the blue negative an image of the sky.
Each of the black-and-white strips yielded, upon printing, a positive relief image in which gradations were represented by varying thicknesses of hardened gelatin. This gelatin absorbed dye of the appropriate color. The dye images were transferred upon coated blank film, one after another, one for each primary color; and the resulting print became the projection copy sent to theatres.
Most color films of the time – with the exception of the self-conscious use of colored lights in the first live-action three-color Technicolor short, La Cucaracha (1934) – in attempting to avoid the charge of “postcard colors,” copied 19th Century naturalistic painting in their exteriors, and the genteel, tasteful designs of home decorating magazines in their interiors. The results were unobtrusive, decorative, but rarely useful in intensifying dramatic effect.
Quite pleasing were the pastel blues and browns and soft skin tones of Magnacolor, a two-color process related to Prizma, Multicolor, and Kellycolor which was first used in Paramount’s short subject series, “Popular Science” (1935-1939) and “Unusual Occupations,” 1937-1939, many of which are represented by original prints in the AFI Collection at the Library. Later, Magnacolor was used in full-length Westerns and melodramas produced by the subsidiary of Consolidated Labs, Republic Pictures.
Kodachrome – introduced in 1935 – had five coatings of emulsion on one side of the film: three layers, each sensitive to its own primary color, and two layers for support. It was used primarily in 16mm non-theatrical productions for schools and industry, as in the 1943 U.S. Department of Agriculture film Tree In a Test Tube (which includes the only color film appearance of the comedy team of Laurel and Hardy).
A later variation called monopack a single-strip color film – was used to record aerial footage for the Technicolor features Dive Bomber (1941) and Captains of the Clouds (1942), films which are among the sixty-five Technicolor features held in their original camera negative form in the AFI Collection. Although results with monopack did not equal those of the three-strip method for Technicolor release, it was the method by which World War II actuality footage in color would be seen in theatres, in such documentaries as With the Marines at Tarawa, Memphis Belle, and Battle of Midway.
Gasparcolor, a three-color print process introduced in the mid-thirties, carried two light-sensitive layers on one side of the film and the third on the other side. Among the most notable examples of Gasparcolor in the Collection are George Pal’s puppet animation advertising film Ship of the Ether, produced for Philips Radio, and the beautiful abstract films of Oskar Fischinger, such as Circles.
Cinecolor, yet another descendant of the Prizma process and perhaps the most successful of all, was first used for a feature film in 1942. In 1948, the combined footage printed by Cinecolor and Magnacolor equaled the total Technicolor footage printed in the United States – more than 200 million feet of two-color film.
For the compromise color reproduction of two-color systems, makeup had to be light to avoid either sallow or red-orange flesh tones; sets and costumes had to be in pastel tones. Cinecolor and an improved version of Magnacolor were capable of producing pleasing skin tones and soft colors, despite an inaccurate tendency toward shades of green and orange. Representative examples in the AFI Collection include a dozen features from the studios of Columbia and Hal Roach, shorts from Warner Brothers and Paramount, including “Popeye” and “Looney Tunes” cartoons. The two-color films survived until the early fifties when Trucolor, a Cinecolor process, became three-color. But even with this improved Trucolor, best exemplified by Nicholas Ray’s Johnny Guitar (1954), Eastmancolor prevailed.
The most important breakthrough since the introduction of three-color Technicolor in the thirties came in 1949 with the debut of Eastmancolor negative and printing stock. Within five years the Technicolor camera was obsolete.
The end of the era of the Technicolor camera coincides with the discontinuation of the use of nitrate film base. Nitrate films are subject to deterioration within forty years of manufacture (a threat hanging over all films, black-and-white and color, on nitrate base, in use prior to the fifties); color films face further dangers, as exemplified by the original negatives for Gone With The Wind: by the late sixties, the base of each of its three-color record negatives had shrunken, each to a different extent, so that it had become impossible to superimpose them in sharp registration by the use of the registration perforations on the films, some of which had been damaged.
Re-registration of the three color records had to be done by eye and required establishing a different relationship among the three positions than that given by the perforations. One color record was established arbitrarily as the standard, and the other two color records were conformed to it.
The color information contained in the black-and-white emulsions in the three-color Technicolor process is relatively stable because black-and-white emulsions are not subject to great change. In contrast, multiple-layer color films, such as Eastmancolor, are subject to some shifts in color and fading of the less stable dyes which provide the color.
Although dye stability has been improved in the last decade, processed film which is to be kept for a long time (for archival purposes or for anticipated reissue) has to be given attention so that color shifts can be minimized. This is done by careful washing at the time of processing, controlling temperature and humidity at the place of storage, and periodic inspection of the color films.
The best method of guaranteeing permanence of the color image is by making black-and-white records, one for each of the primary colors. This doesn’t make a color image immediately available for inspection but a color negative and a print for screening can quickly be made from this negative. It is an expensive method, not only because of the initial preparation of the three separate records, but also for the later reconstitution of the color image, with the subsequent requirements of retiming within a decade’s time.
In the past three-quarters of a century, patent offices have recorded dozens and dozens of ways of effecting color motion pictures. Every possible method – and many an impossible one – has had its moment and then disappeared. Of the few systems that were successful, most are obsolete, their achievements remaining in the two-color and three-color examples which survive.
No color movie process is capable of reproducing every color in the spectrum with complete accuracy at the same time: there is only an indirect relationship between the colors of the subject as perceived by the eye and colors available through dyes. But this is not to deny the effectiveness of Technicolor or Gasparcolor or of the many two-color processes in those films which are their finest achievements. The secret of eloquent color in movies may well lie elsewhere than in an obsession with literal color accuracy.”
(Parker, David L. (1973): ‘Blazing Technicolor,’ ‘Stunning Trucolor,’ and ‘Shocking Eastmancolor.’ In: Tom Shales et al. (ed.): The American Film Heritage, The American Film Institute. Washington, D.C.: Acropolis Books Ltd., pp. 19-27.)
“GLORIOUS AGFACOLOR, BREATHTAKING TOTALVISION AND MONOPHONIC SOUND. COLOUR AND “SCOPE” IN CZECHOSLOVAKIA
The cinema industry was one of the first industries to become state-owned in post-war Czechoslovakia.1 Although state interference in film production, distribution and exhibition grew as the political climate of the cold war became increasingly tense, it did not stop Czechoslovak cinema from following technological changes which were happening abroad. However, isolation from the western world and political and economical dependence on the centre of socialist power in Soviet Russia caused considerable problems. Efforts to evolve independently inside the socialist block were affected by the growing internationalization, standardization and globalization of the cinema industry.
In this chapter I will examine how this tension between Soviet self-sufficiency and a global cinema market affected the adoption of colour in Czechoslovakian cinema in particular in relation to the change of colour process required by the adoption of widescreen. While the Czechoslovak film industry was content to use low-quality East-German colour film stock in the late 1940s, owing to the international adoption of widescreen it was forced to exchange it for Eastmancolor during the following decade. In this period therefore the necessity for technological change powered by the global industry overwhelmed the political realities of the Soviet system.
The first mainstream natural colour films were screened in Czechoslovakia well before 1945, including films utilizing various two-colour systems in the 1920s, American Technicolor productions in the 1930s and German Agfacolor films in the 1940s.2 Although there were minor independent experiments with colour in the period, and Czech workers helped during production of Agfacolor features at Barrandov studios in Prague during the war, regular colour production would start only after 1945.3
The particular character of the Czech film industry between the wars did not allow for the earlier proliferation of colour films. Since the mid 1920s the state would only licence charitable organizations to run cinemas, which discouraged entrepreneurship and meant that film production was not seen as a profitable enterprise and thus was never supported by banks or other private investors. In addition so many distribution companies were set up in the post-First-World-War period, flooding the market with hundreds of films from all over Europe and the U.S. every year, that they did not leave much space for domestic releases, nor did they enjoy long lives in this highly competitive atmosphere themselves (Heiss and Klimeš 2003: 303-320).
After 1945 on the other hand, the state-owned industry was provided financial protection by vertical integration and, on account of the German occupation of the Barrandov studios during the war, not only were experienced workers and fully equipped laboratories available, but also a limited supply of colour film stock. However, before Czechoslovak cinema ventured into its own colour production, it needed more experience, hosting Soviet colour production in the first few post-war years.4 But they could not wait long. Colour production was supposed to prove both the technological and the artistic maturity of the industry. As in other countries, the first attempts at colour cinema were made with short and nonfiction films, and from the latter half of 1945 colour stock was used prominently for both short and feature-length animation. While the focus on short films is understandable due to initial experimenting and high costs of colour stocky the choice of the animation genre not only copied foreign patterns, but also drew on the international reputation of Czechoslovak animation at the time, such as Jiří Trnka’s Animals and Bandits (Zvířátka a Petrovští, 1946) or The Christmas Dream (Vánoční sen, 1945), collaboration of Karel Zeman, Bořivoj Zeman and Hermína Týrlová, both in Festival de Cannes competition in 1946. The young state-owned industry was in need of reorganization and lacked modern equipment in both production and cinemas, as well as the support of domestic manufacturers of technology and film stock.
In such a situation, colour animated films seemed an ideal product to be exchanged for much needed foreign currency.5 For example, in 1947, thirteen out of seventeen short animated films were in colour, as was the only feature-length animated film produced that year, while only one feature out of eighteen and two out of fifty-three non-animated shorts were in colour. At the same time, only some of the films shot in colour were distributed as such at home, the colour copies being reserved for international festivals and the foreign market. The first live action feature film in colour, Jan Roháč of Dubá (Jan Roháč z Dubé, 1947), was made in 1947 and the production of colour films increased steadily every year until the mid-1950s. Even in the critical year of 1951, when only seven feature films were made in total, two of them were in colour.6 The 1950s were also marked by an interesting (but quite understandable) inclination of colour productions towards popular films in general and children’s movies in particular. Children’s and animated films were successful at international festivals and often sold abroad. They constituted prestige product, not only securing the foreign currency, but also showing both possibilities and abilities of the newly nationalized cinema industry, advertising the idea of socialism.
Before 1945, domestic manufacture of film stock was virtually nonexistent, and even in later years only a small amount of black-and-white positive material could be secured internally.7 Czechoslovakia thus depended on foreign supply. The negative colour film stock used well into the 1960s was East German Agfacolor, initially bought through the Soviet Union administration after 1945, and later directly from the Agfa factory in Wolfen (from the mid-sixties, the same stock was called Orwocolor). However, since the mid-1950s, the industry had been experimenting with stock by other European manufacturers and with Kodak products, looking for new and better colour material.
It is important to consider at this point how similar the background for decision-making mechanisms are when it comes to comparison of technological change in the nationalized cinema of a socialist country such as Czechoslovakia, and other cinemas governed by the free market. This is largely due to the nature of cinema as an industry. While in the late 1940s and early 1950s the cinemas of the East-European countries tended towards separatism, as did other industries, quite soon the need of at least partial success in the international market became obvious. Also, while Soviet and other socialist countries’ films were preferred by individual governments, tastes of the audiences in these countries did not differ much from those in the western world (Skopal 2009). Finally, although the main goal of the cinema was to educate the people in the ways of the new and future socialist world, economics constituted an inseparable force behind the control of the industry.
While shooting in colour was not without issues, screening colour films proved to be equally problematic. Firstly, the quality of the eastern Agfa stock was low. In various tests conducted in the period, Czechoslovak technicians found the definition of Agfacolor positive materials 50 per cent lower than that of Eastmancolor, while the sensitivity of the emulsion was uneven, sometimes in the same reel. Up to 10 per cent of the Agfacolor material was sent back to Wolfen as faulty every year. Reports from the period comment on the low quality of the colour stock causing problems during shooting and processing (Anon. 1955). Proof of this is evident in the poor colour saturation in scenes with lower intensity of light (for example night scenes) and changes of colour during dissolves which are visible on the surviving prints and recently released digital copies of some films. Second, domestic cinemas were very poorly equipped for the projection of colour films. Nation-wide surveys showed that some cinemas only had one projector and most of them had machines that were more than twelve years old. Even silent-era equipment, only later adjusted for sound screening, was not unusual. Old projectors were feared to be more likely to damage expensive colour copies during screenings. Furthermore, these projectors had very poor lighting properties. Not only did their optics absorb most of the light before casting it on a screen, but also the light sources were insufficient themselves, as were the reflective qualities of materials used to make screens. Before colour, even a dim projector light was enough: black-and-white films required less light to be sufficiently luminous and cinemas in Czechoslovakia were mostly small, with short distances between projector and screen.8 Screenings in larger venues, however, revealed the inadequacy of the machines.9 Not surprisingly, when reviewing projectors manufactured domestically after 1945, cinema representatives usually complained, about lamp houses and optical arrangement, which had the biggest effect on the light efficiency of the projector.
The survival of Czechoslovak cinema depended on foreign product and the ability to screen films produced abroad.10 As coordination and division of labour and flow of product inside the Council for Mutual Economic Assistance (COMECON) was still poor, Czechoslovakia could not close itself inside the Eastern bloc, at least from the point of view of the cinema market. Being able to screen foreign films and occasionally sell some domestic product abroad was necessary, and therefore if the foreign product was in widescreen, Czechoslovakia needed to be able to adapt to new formats.
The widescreen revolution brought a new set of concerns for Czechoslovak cinema and they were to test the new organization of the industry after little more than a decade of its existence. Firstly, the administration of the centralized cinema industry had to decide which of the emerging new formats to adopt. Having more than one new format alongside academy ratio was impossible for economic and organizational reasons.11 In the initial anarchy of emerging new formats, Czechoslovak cinema technicians had to decide, or rather guess, which of the formats would get the major share of the cinema screens in the world. As they started to consider a new format quite late, around the end of 1954 or the beginning of 1955 (and in these years only preliminary research was made, while the actual adoption was planned for 1956), the chosen format was CinemaScope, which was at that point the dominant widescreen format and had a number of fully compatible competitors. While the word CinemaScope appears (in various distortions of the original spelling) in cinema journals and archival documents of the period, this actual brand never made it to Czechoslovakia, and was substituted by compatible European technology for shooting (for instance French Totalvision), and by domestic equipment for screening.
Now that the decision was made, the next step was to prepare the industry for the transition as quickly as possible. A five-year plan was prepared for the period 1956-1960, during which ten features were supposed to be made, and forty cinemas adapted for the new technology. While the number of films actually made corresponded with the initial plan, over 250 cinemas were adapted during the period. Of these, however, only approximately forty had stereophonic sound reproduction alongside the wider image.
While Agfacolor negative stock was barely sufficient for academy ratio shooting and screening, it was even more inadequate for “scope” or even masked formats.12 As a result, from 1957 the situation became more complicated. For “scope” films, Eastmancolor became the standard, while academy ratio productions continued to use eastern Agfacolor/Orwocolor. We can see this distinction in the film Death in the Saddle (Smrt v sedle, 1958). The film was shot in two formats, academy on Agfacolor and “scope” on Eastmancolor stock.13 A similar situation arose with Provisional Liberty (La Liberté surveillé / V proudech, 1957), the first “scope” feature finished in Czechoslovakia, which was a co-production with the French Trident company. According to negotiations, the French co-producer was supposed to supply the Eastmancolor stock, although in fact what they provided was western Agfacolor. Czech cameramen working on the film found the quality sufficient, although the material needed some changes in lighting and laboratory processing. The release copies were printed on Italian Ferraniacolor.14 A Midsummer Night’s Dream (Sen noci svatojánské, 1959), a “scope” animated feature by Jiří Trnka that went into production in 1956 and was released in 1959, was shot entirely in Eastmancolor. However, in 1956 there was no laboratory to process Eastmancolor in Czechoslovakia, so the rushes were sent to Paris, while Trnka used black-and-white academy materials shot simultaneously to check the movements of his puppets. Also, he was forced to make alterations to his usual methods of puppet-making, due to the differences in colour rendering with Agfacolor and Eastmancolor.15
While the first colour widescreen films were being made, research groups were formed to compare the qualities of Agfacolor and Eastmancolor and to investigate possibilities of introduction of the latter into laboratory practice. The main problems were connected with old and insufficient machinery and inaccurate measuring instruments. However, unlike the Agfa/Wolfen factory, Kodak provided technical support during the introduction of the new laboratory equipment and processes, and remained in contact with the Prague laboratories. While preparing for Eastmancolor as a new negative material, Czechoslovak researchers continued to review other colour stocks produced in Europe. Small groups of technicians (usually two or three) were sent to the Soviet Union, as well as the DEFA studios in Germany, the Gevaert factory in Belgium and the Ferrania factory in Italy. During the 1960s and 1970s, Eastmancolor became and remained the main colour stock for negatives and intermediate materials, while cheaper colour processes (most often eastern Agfacolor, later Orwocolor) were used for distribution copies. There were however a few scope films using Agfacolor or Orwocolor negative film stock, especially in the late 1960s.
Although the widescreen revolution did not take a direct course and was held back, in Czechoslovakia probably more than elsewhere, it did significantly influence the transition from black-and-white to colour. While eastern Agfacolor was perceived as problematic and inferior from the very beginning, the major impulse for a change of colour system came with the introduction of widescreen in Czechoslovakia, as anamorphic processes tested the limits of the film stock, and made all the known issues even more visible. Czechoslovak studios never saw the real CinemaScope and worked with compatible European “scopes”, as the cinemas had to use domestic technology, however clumsy it might be. It was therefore not glorious Technicolor, breathtaking CinemaScope and stereophonic sound at first in Czechoslovakia.
Furthermore, initially Czechoslovak encounters with colour were determined by the character of the cinema industry – low domestic production, small number of cinemas and dependence on international cooperation – rather than by the political situation. Yet after 1945 colour production gained an increasingly prestigious standing from both apolitical and economic point of view. Politically, colour films were meant to show the accomplishments of the state-owned industry, while economically they constituted a unique source of foreign currency. Ultimately the latter half of the 1950s and the 1960s saw a period of development and expansion in the Czechoslovak industry and liberation in politics and society, which encouraged and enabled the government to spend more on the superior Eastmancolor stock. Where colour film stock was concerned, political issues ceased to be important, and once it became economically possible, the cinema technicians and engineers went for the quality first. Eastmancolor was adopted as the standard negative stock material in the following decades, which in turn saw a significant increase in the number of films made in colour, leaving the use of black-and-white marginal by the end of the communist era in 1989.
1 By the government decree no. 50/1945 on arrangements in the cinema sphere of business, the Czechoslovak state gained a monopoly on cinema production, laboratory processing, distribution, public screenings and international trade.
2 Short Prizmacolor films were distributed in 1921-1922. In December 1923 The Glorious Adventure (UK, 1922) premièred and The Toll of the Sea (US, 1922) in two-colour Technicolor was released in December 1924 (Štábla 1982: 355-357).
3 For example IRE-film, a Prague studio owned by Irena and Karel Dodal, made several animation shorts in colour between 1933 and 1938 (Strusková 2006: 99).
4 For example, the shooting of Alexander Ptuschko’s colour film The Stone Flower (Kamennyy tsvetok, USSR 1946) began in August 1945, and the film was the first to be finished in the Barrandov studios after the war.
5 While the studios did not suffer much damage during the war (at least in comparison with other Central and East-European countries), the cinemas were in desperate need of new equipment. According to a post-war survey, up to 85 per cent of them were “insufficient for orderly operation”, not only in terms of equipment, but also in issues of hygiene and safety (Bystřický 1947: unpaged).
6 In 1950, several films were reprimanded by communist party officials and a list of preferred topics was issued. This interference led to a production crisis in 1951, when only a fraction of the originally scheduled 52 films for that year were actually produced.
7 On the other hand, film for still photography had been domestically produced since 1914 by Neobrom, and from 1921 by Fotochema (now Foma), alongside other smaller companies.
8 According to the data collected in 1947, only 110 cinemas had an auditorium longer than 30 metres, and the maximum length was 42 metres (Anon. 1950a: unpaged). In 1950, only nine cinemas had a capacity of more than 1,000, while almost 85 per cent of all cinemas could seat less than 500 (Černík 1954: 89, 198). In later decades, the national cinema network was being improved, also by building new cinemas in previously neglected regions. Of these cinemas, some were constructed especially for widescreen or 70 mm, and as such, they tended to be larger. Also, open-air cinemas, with programming concentrated to summer months, had longer distances between the screen and the projector booth, and could have up to several thousands of spectators. For example, the first “scope” screening in Czechoslovakia during the International Film Festival in Karlovy Vary in summer 1956 took place in a newly constructed open air cinema, which would seat up to 3,500 spectators (Anon. 1956: 9).
9 As the report by a member of Cinema Technology Committee (FITES) states, during the screening at the Fair palace in Prague held in 1950 for the anniversary of the Soviet October Revolution, the visibility of the image was so reduced that a viewer could hardly have recognized that the film was in colour. In conclusion, the report suggested that if the minimum luminance required for colour screening is not achieved, colour films should not be shown at all, as that would ruin their political mission (Anon. 1950b: unpaged).
10 Because of the small number of cinemas, low ticket prices and high costs of production, an average Czechoslovak film would theoretically take thirty-six months to break even, during which time demand would fall dramatically. Thus only extremely popular Czechoslovak films or foreign films turned a profit (Bláha 1955: 12).
11 Although since 1964 some new large cinemas were built for 70 mm projection, Czechoslovak cinema never produced a film on 65 mm negative, except for a handful of co-productions with the USSR. A few other Czechoslovak films were released in 70 mm blown-ups from original 35 mm negatives. It should also be noted that masked formats (1:1,66 and 1:1,85), which could be screened with just small alterations to current projectors, were quite common in Czechoslovakia, but there is no data on their actual proliferation.
12 Only during the 1970s did colour become standard in Czechoslovakia. Until then the majority of both academy and “scope” films were shot in black-and-white.
13 This practice was used for a few early widescreen films, as the Prague laboratories did not have a way to make academy copies from the “scope” originals yet, and at the same time, only a few cinemas were scope-friendly. Having a film on widescreen only would substantially limit its use in distribution.
14 In general, co-production became the way of obtaining quality colour stock and better shooting technology in the late 1950s and during the 1960s (see Skopal 2009).
15 In a monograph on Trnka, the reason offered for switching from Agfacolor to Eastmancolor is the blurriness of the wider image towards the left and right extremes. Also, Eastmancolor is described as more “naturalistic”, showing the materials used to manufacture the puppets for what they really are (Boček 1963: 247-248).
Anon. (1950a) Zápis 60. schůze komise promítací FITES dne 19.6.1950 [manuscript], Filmový technický sbor, FITES 1950, Praha: Národní filmový archiv.
Anon. (1950b) Zápis 65. řádné schůze promítací komise 14. listopadu 1950 [manuscript], Filmový technický sbor, FITES 1950-1951, Praha: Národní filmový archiv.
Anon. (1955) Zápis plenární schůze Filmového technického sboru ze dne 28. září 1955. Zavádění nových technologií. Filmový technický sbor, FITES 1951-1959, Praha: Národní filmový archiv.
Anon. (1956) “Preparations for the IXth International Film Festival at Karlovy Vary in Full Swing”, The Czechoslovak Film, 4: 9.
Bláha, R. (1955) Ekonomika čs. kinematografie. Učebnice pro III. a IV. ročník Průmyslové školy v Čimelicích a příručka pro filmové pracovníky, Praha: Československý státní film.
Boček, J. (1963) Jiří Trnka. Historie díla a jeho tvůrce, Praha: Státní nakladatelství krásné literatury a umění.
Bystřický, J. (1947) Zřizování kin a užití substandardních formátů pro veřejný provoz [manuscript], Filmový technický sbor, FITES 1947, Praha: Národní filmový archiv.
Černík, A. (1954) Výroční zpráva o čs. filmovnictví. Rok 1950, Praha: Československý státní film.
Heiss, G. and Klimeš, I. (2003) Obrazy času / Bilder der Zeit. Český a rakouský film 30. let / Tschechischer und österreichischer Film der 30er Jahre, Praha: Národní filmový archiv.
Pilát, F. (1972) Studie dlouhodobého rozvoje filmové techniky, Praha: Ústřední ředitelství Československého filmu.
Skopal, P. (2009) “The ‘Provisional Liberty’ of Colour and Widescreen: The Czech Co-Productions with the ‘West’, 1959-1969”, paper presented at NECS conference at Lund, 2009. Also online. Available at http://www.phil.muni.cz/dedur/?lang=l&id=21534 (accessed 30 April 2011).
Strusková, E. (2006) “Iréna & Karel Dodal. Průkopníci českého animovaného filmu”, Iluminace, 63: 99-146.
Štábla, Z. (1982) Rozšířené teze k dějinám československé kinematografie, vol. 2, Praha: Filmový ústav.”
(Batistová, Anna (2013): Glorious Agfacolor, Breathtaking Totalvision and Monophonic Sound. Colour and “Scope” in Czechoslovakia. In: Simon Brown, Sarah Street and Liz Watkins (eds.): Color and the Moving Image. History, Theory, Aesthetics, Archive. New York, London: Routledge, pp. 47-55.)
“Compared to Eastman Kodak, Technicolor was a small company. Its success lay in taking financial risks and in investing heavily in research and development. Gorham Kindem compares this aspect of the two companies:
Technicolor’s major inventions involved substantial economic risk. Technicolor continued to invest heavily in research and development despite the fact that between 1923 and 1935 Technicolor, Inc., accumulated a net loss of $2 million. In 1931, Technicolor invested over $180,000 in research and development. Despite the fact that Technicolor possessed a virtual monopoly on three-color for feature films throughout the 1930s, it didn’t actually profit after taxes until 1939.
Eastman Kodak, on the other hand, invested over $ 15 million in color photographic research between 1921 and 1948, but it also had net sales of $435 million compared to Technicolor’s $20 million in 1948. In 1.948, Eastman Kodak also secured its major patents upon colored couplers [substances in the film emulsion or developing solution which form colour dyes during photochemical processing of the exposed film] for Eastman Color, while it invested $3 million in color research. Obviously, Eastman Kodak never undertook financial risks proportionately equal to Technicolor’s in its quest to secure a virtual monopoly over film color through patent protection of its major inventions.16
16 Gorham A. Kindem, ‘Hollywood’s Conversion to Color: The Technological, Economic and Aesthetic Factors’, Journal of the University Film Association, vol. XXXI, no. 2 (Spring 1979) pp. 31-2. ”
(Neale, Steve (1985): The Beginnings of Technicolor. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 13-23, on p. 20.)
“The advent of Eastman Color
The introduction of Eastman Color by Eastman Kodak changed markedly Technicolor’s place within the colour field, leading first to modifications in Technicolor’s practices, technologies and services and eventually, in the 1960s, to a policy of diversification, in which the company expanded its operations to include not only films but television (and not only colour, but also black and white). Eastman Color was a development of the German Agfa-color single-strip process. Its hallmark was that the three strips of colour sensitive film needed for the production of a colour image were bonded together in a single tri-pack roll. Henceforth, a colour film could be shot on an ordinary one-lens camera, while in principle any laboratory with conventional processing facilities could produce colour prints. Technicolor’s monopoly, based on its special camera and on its processing services, was henceforth a thing of the past. Eastman Kodak offered not only negative tri-pack film, but also colour positive and internegative stock as well. Within two or three years, nearly every major studio had adopted Eastman Color negative, while a number used the whole colour series. As the studios adopted and adapted the Eastman Color process, a new series of commercial brand names for colour processes used by the studios began to appear and proliferate: WarnerColor, Ansco Color, TruColor, De Luxe and so on. And this at precisely the point at which television was beginning to have a major impact on the industry, the point at which colour began more and more to be used in film as an attraction vis-à-vis TV’s black-and-white, and the point at which widescreen, 3-D, and other technical developments and novelties were to be used (in conjunction with colour) to intensify the spectacle of cinema in general, the primary means by which the cinema sought to counteract the threat that television posed. An article written by Frederick Foster for The American Cinematographer in 1953, for instance, noted that Eastman Color was being used in particular ‘in the production of many three-dimensional films, where 3-D cameras taking single film, strips are employed instead of Technicolor’s 3-strip cameras’18 and that ‘Twentieth Century-Fox studio is using Eastman color negative in its cameras in the production of CinemaScope films.’19
18 Frederick Foster, ‘Eastman Negative-Positive Color Films for Motion Pictures’, The American Cinematographer, July 1953, p. 332.
19 Foster, ‘Eastman Negative-Positive Color Films for Motion Pictures’, p. 332.”
(Neale, Steve (1985): The Beginnings of Technicolor. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 13-23, on pp. 21-22.)
“From 1953 on, Technicolor would process only Eastman Color negative stock using its peerless imbibition process. Nearly the last to enter the field, Kodak by 1950 had come up with a multi-layered negative stock combining Agfa’s economy and flexibility with Technicolor’s consistency and brilliance. Kodak’s innovation was to eliminate the colourless dye couplers from the emulsion itself and introduce the dyes only in the laboratory. Its original negative stock therefore was essentially three layers of black and white film on a single base mutually self-filtering and recording information about the red, blue and green light entering the lens. In processing, this information was converted into dyes for printing. This could be done conventionally or with the richer, slower imbibition method.”
(Andrew, Dudley (1980): The Post-War Struggle for Colour. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 40-50, on p. 47.)
“4. LES PROCÉDÉS SOUSTRACTIFS
Autour des années trente, on assiste à la naissance presque simultanée des principaux procédés modernes qui relèvent tous du système soustractif. Dans ce système, la couleur est matérialisée sur la copie destinée à la projection. Le quasi-monopole exercé par les procédés soustractifs tient à un fait très banal : il ne demande aucune transformation du matériel de prise de vues et de projection. Mais cette simplification d’utilisation résulte d’une complexité accrue au niveau de la fabrication et du traitement des émulsions. C’est là que se situe le problème du point de vue de la conservation : les trois couches qui composent la structure de la plupart des films soustractifs sont composées de matières colorantes beaucoup plus instables que l’argent réduit de l’émulsion en noir et blanc.
e. Les procédés chromogènes négatif/positif : ces procédés apparaissent en 1941 avec le nouvel Agfacolor dont les brevets seront dispersés après la guerre et donneront naissance à tous les procédés modernes (Gevacolor, Ferraniacolor, Fujicolor, Anscocolor, Sovcolor, Eastmancolor…).
La prise de vues s’effectue sur un négatif comportant trois couches argentiques sensibles respectivement au rouge, au vert et au bleu. Au cours du développement chromogène, des coupleurs ancrés dans chaque couche produisent des colorants cyan, magenta et jaune. L’argent est ensuite entièrement éliminé. On obtient ainsi une image négative où couleurs et valeurs sont complémentaires de celles du sujet. Pour le tirage, on utilise un film du même type qui rétablit les couleurs et les valeurs du sujet. Les procédés chromogènes négatif/positif sont quasiment les seuls utilisés aujourd’hui dans le cinéma commercial. Leur traitement est relativement simple mais cet avantage comporte une contre-partie : l’usine n’est plus à l’extérieur du film comme dans le Technicolor ou le Kodachrome, elle est à l’intérieur. La complexité des multiples réactions qui se produisent dans ce très faible volume fragilise le produit final. Les procédés chromogènes ont été longtemps instables. Les fabricants nous promettent maintenant des émulsions susceptibles de durer plusieurs siècles (sous la réserve du respect des procédures de traitement, d’une bonne utilisation et de conditions de stockage correctes). Il reste néanmoins à sauver quarante années de cinéma en couleur et la tâche n’est pas mince.”
(Pinel, Vincent (1992): La forêt des techniques. In: Michel Ciment (ed.): Ciné mémoire. Colloque international d’information (7-9 octobre 1991). Paris: Femis, pp. 17-24, on pp. 21-24.) (in French)
“There are several color negative films manufactured by different companies throughout the United States and Europe. These negative films can be used in any ordinary black-and-white camera. They have three emulsion layers superimposed on a cellulose acetate base. These three emulsion layers are differently sensitive to different colors of light. This means that the photo-sensitive silver halide particles in the separate emulsions are exposed by different colors of light. Generally, color negative films have a filter layer between the top two emulsions. Where the color sensitivity is not complete, this filter aids in separating unwanted colors from a particular emulsion.
In the United States the most widely used color negatives are Eastman and Ansco.
Both Eastman Color Negative and Ansco Color Negative have only one strip of film surfaced with three layers of emulsion, each being sensitive to a different primary color. Either film can be used in a conventional 35mm camera.
Similar to color negative, 16mm color positive film has three layers of emulsion, each sensitive to a different primary color – red, green and blue, The commercial film is low in contrast and differs from color negative in that a positive color image is obtained by reversal development rather than a negative.
From it three 35mm separation negatives are made when dye transfer release prints are to be made for 35mm exhibition.
In Europe, there are three additional color negatives, Agafcolor, Gevacolor and Ferraniacolor. These negatives are similar to those used in the United States in that three layers of emulsion are superimposed on a single film base.
The Technicolor laboratories, in both United States and England, manufacture release prints from all of these color negative systems.
When photographing with 35mm Eastman or Ansco color negative, any standard 35mm camera may be used, including the “hand-held” or portable models. Specific “color” cameras are not required. After the negative is developed, positive prints may be made in a manner similar to that for black-and-white film, or by the dye transfer method from the color negative, or from separation negatives, as will be explained later.
Only Technicolor offers the producer the alternative of having film printed on color positive stock or by the dye transfer method. Dye transfer release prints offer a cost advantage when a large number of prints are required for worldwide release. And by dye transfer printing from matrices valuable negative is saved from constant re-use.
Color positive release prints are manufactured only from color negative. Color positive stock is similar to color negative in that it has three super-imposed emulsion layers. Color positive stock is contact-printed by light coming through the color negative. Color negative has different colors correlated to the sensitivities of color positive emulsion layers.
Color positive stock records one color image aspect in each of its three emulsion layers and, after printing, is developed.”
(Anonymous (1956): Current Techniques of 35mm Color Film Photography and Printing. In: American Cinematographer, 37,1, January 1956, pp. 26-27 and p. 58.)
Advances in photographic technology during the past few years make possible the incorporation of color-correcting masking in color-photographic processes.1, 2 Color reproduction or masking equations serve as a guide in indicating the kinds and amounts of masking which should be employed. Numerous methods have been suggested for determining optimum sets of equations.3 One of these methods involves the application of the method of least squares to obtain equations for the approximate reproductions of a number of object colors.4 The purpose of the study reported here was to investigate more fully the usefulness of the least-squares method in determining masking equations. Special attention has been given to the effects of (1) the particular selection of object colors, and (2) the assumed set of film spectral sensitivities.
Color-reproduction equations for color-photographic processes may be written in the form of:
c = g10 + g11Dr + g12Dg + g13Db
m = g20 + g21Dr + g22Dg + g23Db (1)
y = g30 + g31Dr + g32Dg + g33Db
The quantities c, m, and y denote the amounts of the cyan, magenta, and yellow dyes in any small area of the photographic film. The dyes are deposited as functions of the exposure densities, Dr, Dg, and Db, of the corresponding small region of the scene photographed. The exposure densities are defined in terms of the exposures R, G, and B as
Dr = -logR Dg = -logG Db = -logB, (2)
R = ∫ SrPTdλ G = ∫ SgPTdλ B = ∫ SbPTdλ. (3)
Each of the quantities inside the integral signs is a function of wavelength, λ; Sr, Sg, and Sb denote the spectral sensitivities of the component emulsions of the film; P denotes the energy distribution of the incident illumination of the scene; and T is the reflectance or transmittance of the small region of the scene under consideration. The energy distribution P is so normalized that for T=1.00 throughout the region of the spectrum to which the film is sensitive, the values of R, G, and B are also equal to 1.00. For this same value of T, Dr = Dg = Db = 0.
The constants g10, g20,and g30 pertain to the color balance of the film. The remaining g‘s indicate the gammas to which the various dyes should be developed as functions of the various exposures. For a chosen set of object colors whose reflectances are known, Eqs. (3) and (2) are applied to determine the exposure densities. The tristimulus values X, Y, and Z for each of the same object colors are also determined. From the known spectral densities of the cyan, magenta, and yellow dyes of the photographic film, values of c, m, and y are determined which yield the same X, Y, and Z values as each of the object colors. For each object color, an equation can then be written which includes the undetermined coefficients g10, g11, g12, and g13, and numerical values for c, Dr, Dg, and Db. If there are more than four object colors involved, all of the equations are not apt to be satisfied by the same set of values for the unknown coefficients. Unique sets of coefficients can be determined, however, by application of the method of least squares to minimize color-reproduction errors (in terms of c) for all the object colors. Similarly, values for the remaining g’s can be found using the same Dr, Dg, and Db values, along with the values of m and y for each of the object colors.
Selection of Object Colors
Ideally, the values of the coefficients in the color-reproduction equations would be relatively independent of the sampling of object colors for which they are derived. Under such conditions, any one set of object colors would be about as good as any other. If, however, the coefficients are markedly dependent upon the set of object colors chosen, these colors should be carefully selected for their importance in the completed color photograph.
Three sets of object colors, of twenty colors each, were selected for this study. The first two sets were chosen on the basis of being important in a color photograph. Specifically, they were selected as follows: A group of engineers and supervisors concerned with quality control in the processing of photographic prints were asked to name the object colors which they considered to be most important in color photographs. Based upon examinations of a large number of color pints and tabulations of the frequency with which various object colors appeared, they provided a list of twenty colors.§ Table I gives this list. Recommended weighting factors were also given, but were not used in the subsequent calculations. Based upon this list, the authors selected two sets of colors, each color representing a particular object corresponding to one of the twenty given in Table I. The chromaticities and visual densities of these two sets of object colors are given in Figs. 1 and 2.
The third set of object colors was selected to span the color space of colors which could be reproduced by the dye set used for the study. The chromaticities and visual densities of this set (Color Set 3) are given in Fig. 3. The spectral densities of unit concentrations of the three dyes from which they were derived are given in Fig. 4. This same dye set was assumed for the color photographic processes for which all the color-reproduction equations were derived.
Six sets of sensitivity distributions were included for investigation. Three of these conform to color-mixture curves. Such choices are required for “exact colorimetric reproduction” in a color-photographic process.5 Each set of color-mixture curves has a corresponding set of primaries. Chromaticities of the primaries are shown in Fig. 5. The three sets of spectral sensitivities are shown as 1, 2, and 3 in Fig. 6. Sensitivities 1 are the colormixture curves, x, y, and z, for the CIE standard observer. They have been renormalized so that, for the assumed illuminant, CIE Illuminant C, the values of R, G, and B in Eqs. (3) are all equal to 1.00 when T = 1.00. Sensitivities 2 are the color-mixture curves corresponding to the monochromatic primaries, R= 620 mμ, G = 530 mμ, and B = 455 mμ. Sensitivities 3 are the colormixture curves corresponding to block-dye primaries, the red primary extending from 585 to 700 mμ, the green from 495 to 585 mμ, and the blue from 400 to 495 mμ.
Sensitivities 2 and 3 contain negative portions which are not found in real processes. Furthermore, even though color-mixture curves are required for exact colorimetric reproduction, they may not provide the best reproduction in processes with available dye systems. Sensitivities 4, 5, and 6, also shown in Fig. 6, were chosen to investigate sensitivities approaching those of real photographic processes. Sensitivities 4 are the chief positive portions (renormalized) of Sensitivities 2. Sensitivities 5 are the chief positive portions of Sensitivities 3. Sensitivities 6 are those of an actual color photographic process.
Color-reproduction or masking equations were determined for each of the three sets of object colors with each of six sets of sensitivity distributions. The coefficients for all the equations (g‘s) are given in Table II. The method of least squares, as described in the introductory section, was employed. Reproduction errors in terms of Δc2, Δm2, and Δy2, each summed for the 20 colors, were minimized. Goodness-of-fit measures, in the form of root-mean-squares of the deviations (or standard deviations or sigmas), are shown under each set of equations as σc, σm, and σy.
Effects of Choice of Object Colors
In Table II the color-reproduction equations for Color Set 1 and Color Set 2 are similar to each other for most of the sets of sensitivities. The sigma values are also of the same general order of magnitude. For Color Set 3, however, almost all the terms in the equations are greater in absolute magnitude than for the other two color sets. Greater masking corrections are called for. Also, the sigma values are somewhat larger, showing that these equations did not give as good reproductions of the colors of Color Set 3. Reference to Figs. 1-3 indicates that the chromaticity and visual density gamuts for Color Sets 1 and 2 are more restricted than are the chromaticity and visual density gamut of Color Set 3. The choice of colors to be used in determining the color-reproduction equations does affect the coefficients in the equations. The effect appears to be associated with the gamut of the colors to be covered. Color Sets 1 and 2 sample more or less the same region of color space, and, even though no pairs of individual objects are exactly alike, give comparable results. The only marked inconsistency between the equations obtained for Color Set 1 and Color Set 2 are for Sensitivities 3. The magenta-dye reproduction equations are:
Color Set 1:
m = 0.007 – 0.169Dr +1.327Dg – 0.180Dg (4a)
Color Set 2:
m = -0.003 – 0.405Dr +1.796Dg – 0.392Dg (4b) The goodness-of-fit values for Color Set 1 were much worse, being 0.068 (σm) as compared to 0.011 for Color Set 2. Examination of the reproduction errors for the individual object colors revealed unusually large errors of 0.137 in the magenta reproduction for Object Color 4 and -0.146 for Object Color 20. Both of these colors are seen to be relatively saturated purples, having high-green exposure densities.
To ascertain the effects that these two colors had on the reproduction equations, they were eliminated from Color Set 1, and a recalculation of the masking equations was made. The equation for the magenta dye then was
m = 0.003 – 0.392Dr +1.740Dg – 0.353Dg (4c)
with σm = 0.011. Reasonably good agreement with the equations for Color Set 2 was thus obtained. Changes in the coefficients in the equations for c and y were not large, but were all in the direction of giving closer agreement with the equations of Color Set 2.
It should be pointed out that Object Color 32, also a saturated purple, was eliminated in the first calculation for the masking equations with Color Set 2 and Sensitivities 3. The elimination was necessary because, for Sensitivities 3, its green exposure value, G, was negative. No value for Dg could, therefore, be assigned to it. The only other colors eliminated from any of the calculations were Object Colors 42, 47, and 48 from Color Set 3. These were also for Sensitivities 3 and for the reason that some of the exposures had negative values.
These results [Eqs. (4a), (4b), and (4c)] again emphasize the importance of sampling the same region of color space if comparable sets of equations are to be obtained with different starting object colors. They also are indicative of the large changes which can be produced in the equation coefficients and goodness-of-fit values by a very small number of colors of high saturation.
Effects of Choice of Sensitivity Distributions
Large, and fairly systematic, differences in coefficient values are obtained for the different sensitivity distributions. For the color-mixture curve sensitivities, there is a downward trend in coefficient value size between Sensitivities 1 and 2 and between Sensitivities 2 and 3. These sensitivities, in the same order, correspond to primaries of decreasing saturation (see Fig. 5). Within the range studied, the more desaturated the primaries, the smaller will be the masking corrections. Sensitivities corresponding to the less saturated primaries, however, have larger and more extensive negative portions. As already indicated, these negative portions (even if obtainable) would have the effect of eliminating highly saturated colors from the gamut of colors for which positive exposures are obtainable. Such colors are outside the range of colors which can be reproduced by positive amounts of the primaries assumed in establishing the sensitivity distributions; they are outside the triangle formed on the chromaticity diagram by the three primaries.
With few exceptions, coefficient values for Sensitivities 4 and 5 are larger than for the corresponding sensitivities which include the negative portions. This seems reasonable to expect because the negative portions of the curves, in effect, perform some of the masking. Their elimination means that greater amounts of masking will be required. Coefficient values for Sensitivities 6 are, in general, smaller than for Sensitivities 4 and 5. This probably results from the greater spectral separation of the red, green, and blue spectral sensitivities for Sensitivities 6. The red, green, and blue exposures are more fully separated, and therefore less masking is required. The coefficient values are larger than for Sensitivities 3, but the negative portions of the curves for Sensitivities 3 have an even greater effect in exposure separation.
Trends in goodness-of-fit values with changes in sensitivities are not readily evident. There is an apparent slight tendency for the sigma values to become smaller for sensitivities corresponding to primaries of decreasing saturation. For the most desaturated primaries (Sensitivities 3), however, some of the more saturated colors must be eliminated from consideration. Even those colors which are near the outer limit of colors for which positive exposure values are obtained can cause decided increases in the sigma values. This effect is both direct and indirect in that, by effectively altering the masking equations, goodness-of-fit values for other colors are made worse.
Elimination of the negative portions of the sensitivity distributions resulted in slightly better goodness-of-fit values in most cases. It is therefore evident that sensitivities which are color-mixture curves, as required for exact colorimetric reproduction, do not necessarily give the best approximations to colorimetric reproduction when exact reproduction is not obtainable. Sensitivities 6 depart materially from color-mixture curves. The spectral regions of their peaks of absorption are considerably displaced from those of any possible set of color-mixture curves. Even so, the color reproductions obtainable with such sensitivities are not materially worse than those for the color-mixture curves giving the best color reproduction. Compared to Sensitivities 1, the only color-mixture curves which are positive throughout the spectrum, Sensitivities 6 appear to give colorimetric reproductions which are definitely better.
Neutral Scale Reproduction
Neutral colors in a scene give red, green, and blue exposure density values which are equal to each other. For a perfectly reflecting, perfectly diffusing white, Dr = Dg = Db = 0. For other members of the series of grays, the exposure density values equal each other but are greater than zero.
Reproduction of the high-intensity white depends only on the values of g10, g20, and g30. For Color Sets 1 and 2 in Table II, the absolute value of most of these quantities is less than 0.01, showing a reproduction which is very nearly neutral. Somewhat larger values are obtained for Color Set 3, but the largest is less than 0.20.
The gamma of the neutral-scale reproduction equals the sum of the coefficients of Dr, Dg, and Db. If nonselective grays are to be reproduced at constant-color balance, the neutral-scale gammas for the c, m, and y equations in each set must all equal each other. If the grays are to be reproduced at their true relative luminances, these gammas will all be equal to 1.00.
Most of the neutral-scale gammas for the natural object sets of equations in Table II differ from unity by less than 0.01; the largest differs by only 0.022. Thus, true relative luminance reproductions of gray scales at essentially constant-color balances are indicated.
Departures from neutral-scale gamma values of unity are slightly greater for Color Set 3. The greatest departure is in the equation for y for Sensitivities 6 where the value is slightly over 1.2. This departure from unity can probably be accounted for as follows:
The twenty colors of Color Set 3 represent different combinations of the dyes of Fig. 4. The neutral for this dye set is somewhat selective, having a rather marked drop-off at the short-wavelength end of the visible spectrum. The blue spectral sensitivity curve for Sensitivities 6 is displaced toward the short-wavelength end of the spectrum as compared to all the other sets of sensitivities. Because of the displacement and the low-density values in this spectral region for neutral colors, the blue-exposure densities will tend to be low in value. To provide adequate yellow dye for reproduction of the neutral scale (and other colors), the coefficient of Db must therefore be unusually large.
The relatively high neutral-scale gamma for the yellow dye of Color Set 3 and Sensitivities 6 is thus seen to be an artifact of the particular set of colors chosen for reproduction. The spectral characteristics of these colors have a consistency of pattern not found among natural object colors. The equations derived for them do not properly apply to natural objects.
Summary and Conclusions
Using the method of least squares, color-reproduction equations for three different sets of object colors for each of six different sets of sensitivity distributions have been determined. For sensitivity distributions which are color-mixture curves, the mask gammas tend to decrease in size with decreasing saturation of the primaries corresponding to the color-mixture curves. Although there also appears to be a slight tendency for goodness-of-fit values to improve with decreasing saturation of the primaries, the results are erratic and we doubt that the differences are significant. We therefore believe that, for a dye set such as used in this investigation, there is no unique set of primaries, or corresponding sensitivity distributions, which is most satisfactory.
Elimination of the negative portions of the sensitivity distributions tends to increase the values of the mask gammas, but not to worsen the goodness of reproduction obtainable. Sensitivities similar to those used in practice, which depart materially from color-mixture curves or their abridgments, give color reproductions essentially as good as any of the other sensitivities studied.
The gamut and nature of the object colors used in determining color-reproduction equations by the least-squares method have a marked influence on the mask values and on the goodness-of-fit measures. Sets of colors with spectral patterns not characteristic of natural objects may give results which are not properly applicable to natural objects. If special spectral patterns are avoided and the color gamut is restricted to the less-saturated colors which appear to have greatest importance in color photography, consistent results are obtained, even though the specific sets of object colors are different. If sensitivities which are color-mixture curves are used, the colors must also be restricted to those well within the gamut which can be matched by mixtures of the primaries corresponding to the color-mixture curves.
1 W. T. Hanson, Jr. and P. W. Vittum, PSA Journal, 13, 94-96 (1947).
2 W. T. Hanson, Jr., J. Opt. Soc. Am. 40, 166-171 (1950).
3 Evans, Hanson, and Brewer, Principles of Color Photography (John Wiley and Sons, Inc., New York, 1953), pp. 639-661.
4 Brewer, Hanson, and Horton, J. Opt. Soc. Am. 39, 924-927 (1949).
5 A. C. Hardy and F. L. Wurzburg, Jr., J, Opt. Soc. Am. 27, 227-240 (1937); see also reference 3, pp. 617-618.
§ The authors are particularly indebted to Mr. John H. Baker, of the Color Print and Processing Department, of the Eastman Kodak Company, for this list.”
(Hanson, Wesley T. (1955): Subtractive color photography. Spectral sensitivities and masks. In: Journal of the Optical Society of America, 45, 1955, pp. 476-481.)
“Red, Blue, Godard
Godard’s first color film was Une femme est une femme (A Woman Is A Woman, 1961); two years later he dealt with color for the second time in Le Mépris (Contempt). In both works the colors are dominantly primaries (“In Le Mépris I was influenced by modern art: straight color, ‘pop’ art. I tried to use only the five principal colors.” – Godard in the New York Film Bulletin [No. 46; 1964], p. 13).
Red and blue are the colors appearing most frequently in both A Woman Is A Woman and Contempt; the recurrence of these hues in a variety of contexts suggests thematic implications. The films are also related in that their primary themes are love triads (a motif which later became geometrically equilateral in The Married Woman); in both, female nudity has the important function of finalizing a precarious relationship. Both are parodies, the former more obvious and comic while the latter is complex, oblique, and tragic.
In each film there is a difference in rhythm which corresponds to the difference of sense. A Woman Is A Woman is quick, choppy, compact and widely varied in locations while Contempt, although thematically complex, is much more slowly paced, has fewer locations and much longer development of individual sequences. Along with these changes in sequential methodology in the latter film, there is a change in the handling of the camera itself. For the most part, fragmented editing is replaced with full-length takes and camera movements are slow, smooth, and calculated. In Contempt, this not only facilitates the tragic sense but is of importance to the work’s visual construction. It is well known that working in color often creates new problems for the intelligent director – an excellent description of these problems was given by Antonioni when he was interviewed by Godard. (See the English edition of Cahiers du Cinema – No. 1, 1966, pp. 28-9.) Godard, through his experience with A Woman Is A Woman, seemed to learn that if color was to function thematically, he would have to extend the length of single shots and slow down his camera movements to allow the viewer adequate time for concentrating on the composition of colors.
Even a simple and incomplete inventory of the recurring colors in A Woman Is A Woman indicates the importance of hue in relation to characterization and narrative development. Angela, the character who motivates the film’s action, is first seen in a red nightclub; her eyelids are shadowed blue. She is shown wearing a white coat and lives in a white apartment with her lover Emile. Camille and Paul, in Contempt, also live in a white apartment. In both cases, the white seems to underscore conditions of neutrality and/or situations whose final outcome is still ambivalent – Angela very much wants a child by Emile, but Emile, who is cool to the idea of Angela having a baby, wears dominantly blue clothing. The neutral ground of the apartment contains a balance of red and blue objects: window awnings, clothespins, drinking cups in the bathroom, a sports poster on the wall in the living room, flashlights, a red lampshade and a blue bedspread. Seen through windows are blue and red neon signs that consistently comment upon the emotional climate of each scene which occurs in the apartment. Angela is also characterized as indecisive at several points; one time she has on one red and one blue stocking, and another time she wears a red and blue plaid dress. There are red dots on her underpants. After being repeatedly refused by Emile, Angela goes to Emile’s friend Albert to conceive. Albert, the film’s straight man, wears grey and feels no real affection for Angela; he is, however, delighted to help her out. At this point Angela is wearing a blue dress and has switched to a black coat. When she returns to Emile, after having intercourse with Albert, she still wears blue and the dots on her underpants are also blue. When she informs Emile, however, the action is still ambivalent and Angela again wears the white coat. The film ends with Angela and Emile in bed, still under a blue blanket; both are sad and confused. Then Angela thinks of a way to solve the dilemma: red neon light pulsates into the apartment and Angela takes off her nightgown for a willing Emile.
Very rarely, since Eisenstein’s Ivan the Terrible, has color in a commercial feature been used except to add a market value. When it has been dealt with at all, it has been used primarily for the enhancement of mood in separate scenes. Godard has attempted a more ambitious function for hue in A Woman Is A Woman: color is used as a leitmotif which parallels and comments upon the narrative theme.
If a color leitmotif is to be used, some system for structuring the colors must be created. In regard to the red and blue motif of A Woman Is A Woman, Kabuki make-up authority Masaru Kobayashi’s comments are important: “. . . the basic colors employed in kumadori are red and blue. Red is warm and attractive. Blue, the opposite, is the color of villains . . .” (The Film Sense, p. 137). These stylized, symbolic color values are more than likely formalizations of direct sensual experience, formalizations based upon relationships of hue sensation and inner emotional states (what Wassily Kandinsky called “der innere Klang”). Eisenstein felt that these alleged correspondences of sensation and emotion could not be the basis for the systematic organization of color due to the high degree of variation in subjective responses persons have to hues; instead, he suggested that each film create its own “functional” system of organization, using arbitrarily chosen but consistently recurring colors or values. Godard’s color system is in accord with Eisenstein’s theory insomuch as it is “functional” and its colors do not act upon the viewer in a directly sensual way. Godard admitted this himself when he made the following comment about a film in which each composition (through filtering and juxtaposition of hues) creates color “auras” that establish emotional responses in its viewers: “I was very impressed with the new Antonioni, The Red Desert: the color in it was completely different from what I have done: in Le Mépris the color was before the camera but in his film, it was inside the camera.” (New York Film Bulletin, loc. cit.) On the other hand, Godard’s dominant thematic hues were very likely not chosen arbitrarily since they have such obvious symbolic references to emotional states.”
(Sharits, Paul (1966): Red, Blue, Godard. In: Film Quarterly, 19,4, 1966, pp. 24-29, on pp. 24-26.)
“Gegenstand dieser Abhandlung ist eine Methode zur mathematischen Nachahmung des ganzen Farbwiedergabevorganges, so daß an jeder beliebigen Stelle Änderungen von Parametern in die verschiedenen Stufen dieses Verfahrens eingeführt werden können. Dadurch werden gewisse Möglichkeiten für die Prüfung der Farbwiedergabequalität von Farbmaterial geschaffen, was bisher in empirischer Weise noch nicht mit der erforderlichen Genauigkeit möglich war.
1. Mathematische Methode
Das Farbwiedergabeverfahren wird in elementare Stufen aufgeteilt, und jede Stufe wird in eine für die mathematische Behandlung leicht zugängliche Form übersetzt. Abb. 1 gibt das Blockschema des Rechenprogramms. Änderungsfähige Stellen sind umkreist.
Eine Reihe von 51 Farbvorlagen (Abb. 2), gleichmäßig über die Farbtafel verteilt, wird als Standard-Aufnahmeobjekt gewählt. Dazu gehören Konzentrationsreihen derselben Farbstoffe. Als repräsentativ wurden dunkle und helle Farben, sowie auch die mittlere menschliche Hautfarbe, das Himmelsblau und das Blattgrün (Evans1 S. 493-497, Brenman2) genommen.
Ausgehend von den spektralen Remissionskurven dieser Muster wird die Aufnahmedichte für eine gegebene Spektralempfindlichkeit des Negativtyps sowohl für Umkehr- als auch für Negativmaterialien berechnet.
Die Belichtungsdichtewerte werden in äquivalente Graudichtewerte der Kopie unter Verwendung einer standardsensitometrischen Kurve umgesetzt. Ein Vergleich von Wiedergabesystemen mit verschiedener Tonskala ist nur mit einem Bezugspunkt möglich. Wir wählten den Wert 0,64 als genau wiederzugebende Graudichte, weil in der Praxis festgestellt worden ist, daß dieser Wert am besten der Originaldichte gleichkommt.
Annähernd können die charakteristischen Kurven eines Negativmaterials als Gerade angenommen werden. In diesem Fall weiden die Belichtungsdichten in Maximaldichten der Negativfarbstoffe umgesetzt. Ausgehend von diesen 3 × 51 Maximaldichtewerten werden die spektralen Durchlässigkeitskurven der 51 Negative berechnet unter Verwendung der spektralen Durchlässigkeitskurven der Negativfarbstoffe, einschließlich der Maskenfarben, falls anwesend. Von diesen 51 Negativen berechnen wir die Kopierdichten für eine gegebene positive Spektralempfindlichkeit. Diese Kopierdichten werden an Hand einer positiven sensitometrischen Kurve in äquivalente Graudichten umgesetzt.
Zwei Bedingungen sind bei dieser Berechnung vorausgesetzt:
a) Die Originalgraufarbe der Dichte 0,64 wird, wie im Umkehrverfahren, gleich wiedergegeben.
b) Die Originalgraufarbe der Dichte 1,56 wird als neutrales Grau, im allgemeinen aber nicht mit derselben Dichte wiedergegeben.
Die Bedingung a) wird durch eine relative Verschiebung der drei Positivkurven erfüllt, die Bedingung b) durch Änderung des Gammas der gelben und blaugrünen Kurve bezüglich der purpurnen Kurve.
Von diesem Punkt an verlaufen die Umkehr- und Negativ-positiv-Programme gleich.
Die drei äquivalenten Dichten der 51 Vorlagen werden in die Konzentrationen der gewählten Positivfarbstoffe umgesetzt. Jeder Konzentration entspricht eine bestimmte spektralphotometrische Kurve; die subtraktiven Kombinationen der 3 × 51 spektrophotometrischen Kurven ergeben die 51 Kurven der Wiedergabefarben. Daraus werden die Normfarbwerte für Tageslicht berechnet. Um ein quantitatives Maß der Unterschiede zwischen Original- und Wiedergabefarben zu bekommen, werden die Normfarbwertanteile x, y in die α-β-Koordinaten des gleichmäßigen Scofield-Judd-Hunter-Diagramms3.
Schließlich erhalten wir die Länge des Raumvektors Δ in NBS-Einheiten durch Verwendung der Judd-Hunter-Formel [1 NBS-Einheit = 5 J.P.S. (just perceptible steps): gerade noch wahrnehmbare Stufen]4. Für jede Farbengruppe eines bestimmten Farbtons wird die mittlere Länge des Raumvektors berechnet und tabelliert. Zum Schluß wird ein Gesamtfarbwiedergabekoeffizient berechnet, wobei jeder Farbgruppe einem Gewichtsfaktor zuerkannt wird, je nach dem von einem durchschnittlichen Beobachter darauf gelegten Wert. Die Berechnungen wurden auf einem Wegematic 1OOO-Rechner durchgeführt, einer schwedischen Ausführung des Alwac III E-Rechners.
Für die verschiedenen Veränderlichen des Systems wurden Daten von aktuellen Farbmaterialien abgeleitet, nämlich
zwei negative sensitometrische Kurven,
drei Sätze Negativfarbstoffe, von denen zwei maskiert sind,
eine große Menge Positivfarbstoffe und ihre Kombinationen.
Die Gewichtsfaktoren der verschiedenen Farben wurden durch ein Referendum unter 150 Personen bestimmt. Als Gammawert, auf den wir uns in diesem Beitrag beziehen, wird der Richtungskoeffizient der Verbindungsgeraden zwischen dem Punkt 0,70 über dem Schleier und dem um 0,45 log (I · t) weiter liegenden Dichtewert definiert.
3.1: Einfluß des Gesamtgammas
Abb. 3 zeigt den Einfluß des Gesamtgammas. Diese Kurve wird durch die Berechnung für das Negativ-Positiv-System durch alleiniges Variieren des Gammawertes des Positivmaterials erhalten. Der mittlere Farbunterschied zwischen Original- und Wiedergabefarben ist als Ordinate, das Gesamtgamma als Abszisse aufgetragen.
Die Farbwiedergabe verbessert sich mit steigendem Gamma (der mittlere Farbunterschied wird kleiner), bis ein optimaler Wert bei γ = 3,25 erreicht wird, wo die mittlere Differenz Δ = 22,8 NBS-Einheiten beträgt. Jedoch erreichen nicht alle Farben die beste Wiedergabe für denselben Gammawert (Abb. 4): Blattgrün bei γ = 1,90, Rot bei γ = 2,40, Gelb und Blau bei γ = 3,20, während die Farbe der menschlichen Haut und das Himmelsblau nicht einmal die beste Wiedergabe für γ = 4,00 erreichen.
Der Gesamtwiedergabekoeffizient zeigt also einen optimalen Wert in einem breiten Gammabereich.
Die Erklärung, weshalb Rot und Blattgrün ihren minimalen Δ-Wert bei niederen Gammas erreichen, kann in der Farbtafel der 51 Wiedergabefarben gefunden werden (Abb. 5).
Für ein relativ niedriges Gamma (z. B. γ = 1) werden alle Farben weniger gesättigt als ihre Originale wiedergegeben. Mögliche Gründe dafür sind:
a) die unvollkommene Spektralempfindlichkeit des negativen Materials;
b) die unvollkommene Maskierung der Negativfarbstoffe;
c) die unvollkommene Abstimmung der spektralen Empfindlichkeit des Positivmaterials auf die Absorption der Negativfarbstoffe;
d) die spektralen Absorptionseigenschaften der Positivfarbstoffe.
Durch Erhöhen des Gammas des Positivmaterials werden alle Farben gesättigter und näher bei den ursprünglichen Farben wiedergegeben (Abb. 6). Die in den steilsten Bereich der Positivkurve fallenden Farben werden zuerst die beste Wiedergabe erreichen. Die in den unteren Teil der Kurve fallenden Farben werden ein hohes Gamma für die optimale Wiedergabe erfordern.
Eine Farbe der ersten Sorte ist Blattgrün (Sommerfarbe), dessen Belichtungsdichten im Rot, Grün und Blau alle über D = 0,73 liegen, so daß die entsprechenden Negativdichten niedrig sind und im steilen Gebiet der Positivkurve kopieren.
Dieselbe Überlegung gilt für die gesättigste rote Farbe (Abb. 7) (Dichte Rot = 1,92, Grün = 1,79, Blau = 0,53); das mittlere Rot aber erreicht seine optimale Wiedergabe nur für höhere Gammawerte, weil den niedrigeren Konzentrationen auch Rechnung getragen wird. Die niedrigsten Konzentrationen fallen tatsächlich in den unteren Teil der Kurve und benötigen ein hohes Gamma für bessere Wiedergabe. Dies ist auch der Fall für das Himmelsblau und die Farbe der menschlichen Haut. Dieser Gedankengang wird bestätigt, wenn man eine Gerade als positive charakteristische Kurve verwendet. Die Farbwiedergabe mit solchem Material ist viel besser als mit einem Material mit S-förmiger Kurve in zwei Beziehungen:
a) dieselbe Qualität der Farbwiedergabe wird mit weniger exzentrischen Gammawerten erhalten;
b) die beste erzielbare Wiedergabe ist besser als die beste erzielbare Wiedergabe mittels eines Materials mit S-förmiger Kurve.
Tatsächlich erreichen mehr Farben ihre minimale Länge des Raumvektors für dasselbe Gamma (Abb. 8), so daß das Gesamtminimum enger und niedriger wird (Δ = 19,25 NBS). Das Minimum liegt bei niedrigerem Gammawert, weil das Kopiergamma nun für alle Farben, wenigstens im niederen Gammabereich, dasselbe ist.
Wie oben schon erwähnt, fallen die beste Wiedergabe der ungesättigten Farben und das Mittel der diese Farben enthaltenden Farbgruppen noch auf die rechte Seite des allgemeinen Minimums. Die niedrigsten Dichteweite dieser Farben werden bereits bei niedrigem Gamma durch die Dichte Null wiedergegeben, und die anderen Dichtewerte unter 0,64 werden durch steigendes Gamma herabgesetzt, so daß die Wiedergabe schlechter wird.
Purpur 1, Grün 1 und Rot 1 (Abb. 9) sind typische Farben, welche über γ = 1,34 als Weiß wiedergegeben werden.
Diese Betrachtungen gelten auch für das Umkehrsystem, andere Negativkurven, Negativfarbstoffe, Maskierfarbstoffe, positive Spektralempfindlichkeiten und positive Farbstoffe, unter der Bedingung jedoch, daß die quantitativen Resultate in jedem Einzelfall verschieden sein werden.
3.2: Einfluß des Negativ- und Positivgammas für ein konstantes Gesamtgamma
Für eine Reihe von negativen Geraden mit verschiedener Neigung, aber mit derselben Grauabstimmung haben wir das Positiv angepaßt, um ein konstantes Gesamtgamma zu behalten. Abb. 10 gibt die Resultate bei zwei Gesamtgammas (1,98 und 2,62).
Für Negativgammawerte zwischen 0,5 und 0,8 sind die Differenzen in der Farbwiedergabe so klein, daß sie nicht ins Gewicht fallen. Für höhere Negativgammas (γ > 1) werden die Resultate schlechter.
Diese Tatsache ist folgendermaßen zu erklären: Die Kopierdichte des Negativs nimmt mit steigenden Konzentrationen an Negativfarbstoff und demnach mit steigendem Negativgamma zu. Diese Beziehung ist aber nur linear in nahezu grauen Wiedergaben für kleine und für höhere Negativdichten. Für höhere Negativkonzentrationen ist die Beziehung für gesättigte Farben nicht mehr linear, und die Nichtlinearität nimmt mit steigender Sättigung zu.
Wenn also andere als farbmetrisehe Erfordernisse ein bestimmtes Negativ- oder Positivgamma wünschenswert machen (z. B. wegen des Belichtungsspielraumes bei der Aufnahme oder beim Kopieren von Negativen bei gleichbleibendem Gesamtgamma), wird die Farbwiedergabe, wenigstens im praktisch verwendeten Negativgammabereich, nicht erwähnenswert beeinträchtigt.
3.3: Das Umkehrverfahren
Der Einfluß des Gammas und der Kurvenform auf die Umkehrfarbenwiedergabe ist im allgemeinen derselbe wie für das Negativ-positiv-System. Es ist jedoch nützlich, beide Systeme zu vergleichen, weil sie einige typische Differenzen aufweisen. Das Umkehrverfahren ist vergleichbar mit einem eine völlig maskierte Stufe enthaltenden Negativ-positiv-System.
In Abb. 11 werden verglichen: die Wiedergabe mit einem unmaskierten Negativ, die Wiedergabe mit einem völlig maskierten Negativ, jene mit einem praktischen Negativ (einigermaßen übermaskiert in der blaugrünen Schicht) und die Wiedergabe mittels des Umkehrverfahrens.
In jedem einzelnen Fall wurden die Berechnungen für dieselbe Aufnahmespektralempfindlichkeit, dieselbe Positivkurve und dieselben Positivfarbstoffe gemacht. Wenn man die Resultate für das praktische Wiedergabegamma 2,1 vergleicht, wurde ein durch völlige Maskierung herbeigeführter durchschnittlicher Gewinn von 28,8 – 27,8 = 1 NBS-Einheit in der Farbwiedergabe erhalten. Mit einer Übermaskierung der blaugrünen Schicht, wie im Falle 3, wurde ein Gewinn von 28,8 – 26,2 = 2,6 NBS-Einheiten gefunden.
Beim Betrachten der Wiedergabe mit dem Umkehrsystem, bei dem dieselbe Gesamtwiedergabekurve verwendet wird, stellen wir eine Verbesserung von 28,8 – 22,7 = 6,1 NBS-Einheiten fest. Dies bedeutet eine Verbesserung, die sechsmal größer ist als diejenige mit völliger Maskierung. Unter Gleichhaltung aller anderen Elemente kann dies nur durch die ungenügende Abstimmung der Negativfarbstoffe auf die Positivspektralempfindlichkeit erklärt werden. Wenn z. B. ein Negativfarbstoff nicht völlig die positive Sensibilisierung deckt, ist die Kopierdichte nicht proportional der Konzentration des Negativfarbstoffes. Dies kann vorkommen bei Verwendung von Negativfarbstoffen mit zu engem Spektralbereich, von zu breiten Positivspektralempfindlichkeiten und besonders von zu hoher Blauempfindlichkeit der rot- und grünempfindlichen Schichten. Es ist immer möglich, die Abstimmung mit Farbfiltern auf Kosten der Lichtintensität zu verbessern.
3.4: Einfluß der Maske
Wenn wir die Wiedergabe des aktuellen maskierten Negativs im einzelnen betrachten, finden wir, daß der Gewinn für Blau, Gelb, Rot und Purpur (Abb. 12) am größten ist. Für die übrigen Farben ist der Einfluß weniger ausgeprägt; dennoch wird die Farbwiedergabe für jede dieser Farben verbessert. Die durch Maskierung der blaugrünen Schicht herbeigeführte Verbesserung in der Farbwiedergabe ist mit der mit einer 15%-igen Erhöhung im Gesamtgamma erreichten Verbesserung vergleichbar. Der Vorteil einer Maskierung vermindert sich mit steigendem Gamma, weil die Farbwiedergabe schon für hohe Gammawerte gut ist, so daß Übersättigung stattfindet.
3.5: Einfluß der Farbstoffe des Positvmaterials
Bei einem bestimmten Gammawert (γ = 2,1) wurde mit Hilfe des Umkehrprogramms eine große Menge von Farbstoffen (Gelb, Purpur und Blaugrün) des Positivmaterials zur Bildung eines Dreischichtenmaterials in Reihen zu je drei zusammengenommen. Die Farbwiedergäbe hängt stark von der gewählten Kombination ab und kann in der Praxis von Δ = 20 bis 25 NBS-Einheiten variieren. Eine solche Untersuchung wurde schon von MacAdam5 und Evans (S. 531)1 beschrieben.
Das Betrachten einer großen Menge von Kombinationen erlaubte uns, mehr quantitative Angaben über die Größe der Verschiebungen in der Farbwiedergabe zu sammeln, welche auf die Änderungen in den spektralen Charakteristiken der Farbstoffe zurückzuführen sind. Beim Teilen des Spektrums in drei Gebiete für die Bestimmung von Haupt- und Nebenabsorptionen wurden die Normspektralwerte verwendet, weil ein Positivmaterial visuell beurteilt wird und wir auf diese Weise u. E. möglichst gut die spektralen Charakteristiken des Auges in die Berechnung einbeziehen.
Für eine gute Farbwiedergabe soll ein Gelbfarbstoff möglichst wenig im grünen und roten Gebiet des Spektrums absorbieren und eine möglichst zweckmäßige Blauabsorption aufweisen. Tabelle 1 gibt einen Überblick der Wiedergabequalität einiger Gelbfarbstoffe in Kombination mit einem festen Satz von Purpur-Farbstoffen (p1) und Blaugrün-Farbstoffen (bg3), sowie auch über die auf eine feste Hauptabsorption reduzierten Nebenabsorptionen.
Die hohe Absorption im roten Gebiet braucht uns nicht zu beunruhigen; sie ist eine Folge der speziellen Form der Normspektral wertkurve x̅(λ). Es sei bemerkt, daß sich die Farbwiedergabe verbessert, wenn die Nebenabsorption im grünen Gebiet kleiner wird. Den Sinn und die Größe in NBS-Einheiten der Verschiebung der verschiedenen Farben findet man in der Tab. 2.
Ein weiterer Vergleich der Gelbfarbstoffe führt zu dem Schluß, daß weniger die mit der Maximumdichte übereinstimmende Wellenlänge, als vielmehr die Flankenwellenlänge wichtig ist. g3 und g4 haben fast dieselbe Nebenabsorption im Grünen, während die mit ihrer Maximumdichte übereinstimmenden Wellenlängen weit auseinanderliegen, nämlich bei 460 nm und 428 nm. Jedoch sind ihre Wiedergabekoeffizienten wenig verschieden: 21,46 gegenüber 21,61.
Diese Betrachtungen wurden für die Kombination verschiedener Gelbfarbstoffe mit einem bestimmten Satz von Purpur- und Blaugrünfarbstoffen gemacht. Sie gelten im allgemeinen aber auch für die Kombination mit einem anderen Satz von Purpur- und Blaugrünfarbstoffen.
Prinzipiell ist der beste Purpurfarbstoff ein solcher, der eine hohe Grünabsorption und eine minimale Absorption im blauen und roten Gebiet des Spektrums besitzt. Weil sich die roten und grünen Spektralwertkurven x̅(λ) und ȳ(λ) einander stark überlappen, wird ein zweckmäßiger Purpurfarbstoff noch eine ziemlich hohe Rotabsorption aufweisen. Es muß aber ein Unterschied gemacht werden zwischen der Absorption im blauen und roten Gebiet wegen der Breite des Absorptionsbandes einerseits und derjenigen, die durch eine horizontale Absorption verursacht ist.
Tab. 3 gibt die Wiedergabekoeffizienten für verschiedene Purpurfarbstoffe, kombiniert mit einem gegebenen Gelbfarbstoff (g3) und einem Blaugrünfarbstoff (bg2).
Aus dieser Tabelle geht hervor, daß es nicht genügt, eine niedrige Nebenabsorption im roten und blauen Gebiet zu haben; vielmehr ist die Form der Hauptabsorption sehr wichtig für die zweckmäßige Wirkung des Farbstoffes. Abb. 13 zeigt, weshalb p1, ungeachtet seiner höheren Blau-und Rotabsorption, doch eine bessere Farbwiedergabe ermöglicht als p2; p1nähert sich mehr die Blockform; p5mit einer steilen Flanke an der roten Seite weist an der blauen Seite eine zu große Lücke auf, wodurch ihre Grünabsorption in diesem Gebiet wenig zweckmäßig ist.
Die Tab. 4 deutet die Art und die Größe der durch eine Änderung in den Absorptionscharakteristiken der Farbstoffe herbeigeführten Verschiebungen an.
Wir möchten hier abermals betonen, daß ein Unterschied gemacht werden muß zwischen den durch eine graue Absorption verursachten Nebenabsorptionen und den Nebenabsorptionen, welche die Folge des Verschiebens der Kurvenflanke sind.
Die Qualität eines guten Purpurfarbstoffes ist hauptsächlich seiner niedrigen Graunebenabsorption zuzuschreiben. Die guten Purpurfarbstoffe haben noch eine ziemlich hohe Blauabsorption, weil ihre Flanke stark ins blaue Gebiet durchdringt.
Weil die Normspektralwerte x̅(λ) und ȳ(λ) einander stark überlappen, wird eine gute Rotabsorption immer mit einer hohen Grünabsorption verbunden sein. Die Tab. 5 gibt die Wiedergabekoeffizienten einiger Blaugrünfarbstoffe in Kombination mit g1 und p2 sowie ihre Absorptionseigenschaften.
Durch die starke Überlappung der grünen und roten Normspektralwerte braucht eine hohe Grünabsorption nicht unbedingt ungünstig zu sein. Eine hohe Blauabsorption dagegen ist ungünstig.
Richtung und Größe der Änderungen der individuellen Farben findet man in der Tab. 6.
Wir sind uns darüber im klaren, daß neben den farbmetrischen Erfordernissen noch viele andere Faktoren beim Farbfilm zu berücksichtigen sind. Es war aber unsere Absicht, hier die relative Wichtigkeit der verschiedenen, die Farbwiedergabe beeinflussenden Parameter darzustellen.
1 Evans R. A., W.T. Hanson jr. und W.L. Brewer, Principles of Colour Photography. New York and London: 1953
2 Brenman, E.J., Phot. Sei. Eng. 1 (1957), S. 74
3 Scofield, F., D.B. Judd und R.S. Hunter, A Proposed Method of Designating Color. ASTM-Bull. Nr. 110 (1941), S. 19-24
4 Hunter, R. S., Photoelectric Tristimulus Eolorimetry with Three Filters. J. opt Soc. Amer. 32 (1942), S. 509-538
5 MacAdam, D. L., Colorimetric Analysis of Dye Mixtures. J. opt Soc. Amer. 39 (1949)), S. 22-30″
(Smits, J.; Corluy, H.; De Kerf, J. (1966): Farbmetrische Analyse fotografischer Farbwiedergabe-Verfahren. In: Die Farbe, 15, 1966, pp. 102-118.) (in German)
“A Method of Pre-Exposing Color Negative for Subtle Effect
By Freddie Young, BSC
Speaking to Mr. Dennis Wratten of Kodak Ltd. London recently about a method of pre-exposing Eastman color negative to obtain a subdued color effect, I promised a full report on the procedure so that other lighting Directors of Photography may have all the facts.
I illuminated a dead white card with incandescent lighting of 3200 Kelvin; then, taking an exposure meter reflection reading, working at 50 A.S.A. for a full normal exposure if the camera shutter was 170 degrees on a Mitchell camera, I closed the shutter down to 60 degrees opening; this gave me roughly 30 per cent exposure to normal.
After exposing the film I did a series of tests, shooting normal exteriors and also various types of studio lighting. Incidentally, I also tried pre-exposures of 10 per cent, 20 per cent and 40 per cent, but came to the conclusion that 30 per cent was the most satisfactory for all general purposes.
Of course I immediately realised that this method of exposing the film through the camera with the mask line showing on each frame, together with the danger of camera scratches, etc. would not be satisfactory, so I asked Mr. George Gunn of Technicolor Ltd. to match my 30 per cent pre-exposing in the laboratory without the mask lines showing. This was done and we had Technicolor pre-expose batches of 30,000 ft. at a time, always keeping a week’s work ahead of us.
The whole film The Deadly Affair (Director Sidney Lumet) was shot using this method.
Mr. Lumet, whom I met for the first time on a special visit to New York, had asked me if I could think of some way to shoot the picture with subdued colour. He was delighted with the results obtained as were the Art Director, Make-Up Artist and the actors who viewed rushes.
Since then, Mr. Lumet has screened rushes for several Directors of Columbia Pictures Corporation, Mr. Frankovich of New York included, and they have all been most impressed.
The picture is a mood spy drama, but I would venture to suggest that almost any picture, or certainly sequences, could be photographed using this method to advantage, and the following important points should be noted:
– There is no loss in definition.
– The film speed is increased from A.S.A. 50 with incandescent light to at least 75 A.S.A. and 100 A.S.A. for night effect, all printing normal 15 to 17 printer light at Technicolor Ltd.
– Very little filler light is required in the studio.
– No retakes were experienced in any way, so no snags.
The Deadly Affair was shot in eight weeks, a week under original schedule. Technicolor is now in the process of making various tests such as duping, release printing etc. to endeavour to find out if there are any possible snags to the process.
It seems to me to be the answer to so many Directors’ requests in recent years for subdued colour. David Lean asked me on Lawrence of Arabia for such an effect on the long desert crossing to Akaba, and at that time I used slight fog filtering, but this present method could have been much superior.
To conclude – this method does not change the colour values in any way, it merely softens or subdues colour. The increase in speed of the film is very valuable, and personally I feel the colour is more true to life generally speaking”
(Young, Freddie (1966): A method of pre-exposing color negative for subtle effect. In: American Cinematographer, 47,8, Aug. 1966, p. 537.)
“Remember the Glorious Color of films gone by? It may soon be only a memory – for the prints are fading fast.
The house lights dim, a hush comes over the expectant audience, and a beam from the projection booth hits the screen. But the old film, so clear in a buff’s memory, looks to have deteriorated before his eyes – and memory is not at fault. What was once a color film is now a jarring mixture of faded dyes in a spectrum that runs from dull, muddy pink to deep, garish purple. The sunny, windswept fields of Oklahoma! have turned an eerie, strident pink. Marilyn Monroe looks jaundiced. The florid gold and pastel palace in The King and I is now a drab, dusky rose.
Filmgoers who regularly attend repertory theaters or museum retrospectives are finally getting visual proof that color films will not always appear as they did decades ago, years ago – sometimes even months ago. These color fading problems are not isolated nightmares confined to a few prints of certain films. Color fading threatens all color films, and there is a growing awareness that it has not only reached epidemic proportions but has surpassed all other problems of film preservation. Film companies are scrambling to save their precious libraries, motion picture archivists are watching their collections change hue, and audiences everywhere are subjected to a special visual agony that humane people wouldn’t wish on laboratory mice.
Color fading, if left unchecked, could very well do irreparable damage to film preservation and scholarship. It’s the disease that doesn’t discriminate: it can affect Gone With the Wind as easily as it can The Texas Chainsaw Massacre. Already a litany of horror stories is intoned by film archivists throughout the world: the original negative to Kinugasa’s Gate of Hell is so badly faded that no viewable prints can be struck from it; one reel of the original negative of the 1956 Oscarwinning Around the World in 80 Days is missing an important color, and when the film was reissued a duplicate negative had to be made from a print in good condition; the original negative to The King and I is “shot.” Films that most people think are safely tucked away in some kind of Hollywood Heaven are, in reality, being damaged beyond repair.
Specialists today estimate that the average color print has a life of anywhere from six months to twenty years. Even more alarmingly, the original negatives of many films produced from the Fifties on are also in danger of fading. Needless to say, once the negative goes there is very little left. Already certain distributors have resorted to all kinds of visual “enhancement” in the case of negative fading, but the results are far from satisfactory.
It is not surprising that examples of color fading in film are rampant today. The entire history of full color in motion pictures spans slightly over forty-five years, and flaws in the system are becoming distressingly evident.
From the days of Edison, audiences have craved color in their films. Processes such as tinting and toning enjoyed immense popularity at the turn of the century. Hand-painted prints of The Great Train Robbery (1903) would surprise audiences with red-tinted gunshot blasts, and in C.B. De Mille’s Joan the Woman (1917) the execution sequence featured flames tinged with red. Handcoloring reached its peak in France, where Charles Pathé employed over 300 girls from the French countryside to hand-stencil films, much in the manner of penitent nuns making lace.
But neither tinting nor toning could produce a natural color spectrum on film, and research was concurrently carried on by the Technicolor Company and Eastman Kodak. In 1929 and 1930, Technicolor developed a two-color system which employed orange-red and blue-green dyes. The process, although limited in reproductive quality, was immensely popular, and audiences flocked to such two-color films as Warner Brothers’ On With the Show (1929) and the Florenz Ziegfeld-Samuel Goldwyn production Whoopee (1930).
Reacting to public favor, Technicolor rapidly sought to perfect its color process, and in 1932 developed a three-color system first used in Walt Disney’s animated short Flowers and Trees (1932). Disney thought so much of the three-color process that he used it for The Three Little Pigs (1933) and eventually for his first animated feature, Snow White and the Seven Dwarfs (1937).
The three-color system was used with equally positive results on live-action films. The process, called “three strip,” involved the simultaneous exposure of three separate rolls of color-sensitive film (the three dyes involved were magenta, cyan, and yellow) in a large, bulky camera. Becky Sharp (1935) became the first film to utilize full Technicolor, followed by A Star Is Born (1937), The Adventures of Robin Hood (1938), and Gone With the Wind (1939).
As Technicolor became firmly entrenched in the Hollywood of the Forties, it began to acquire a familiar look. Down Argentine Way (1940) featured a more vibrant Technicolor with bold primary colors, notably a saturated red. The words “In Technicolor” became synonymous with the best of Hollywood’s musicals and epics.
In the early Fifties, however, an event occurred in color technology that would revolutionize color film. Eastman Kodak announced the development of a “multi-layer” film which already contained the three dye-sensitive layers necessary for color. This not only did away with the need for the bulky three-strip cameras, but also changed Hollywood’s method of striking prints. Until then, Technicolor struck prints utilizing the “imbibition” process, in which the three different color dyes were each applied separately to the film base, much in the manner of color lithography. But with Eastman’s new “multi-layer” film, colors could be obtained quicker and easier. Although the Eastmancolor prints were inferior in quality to Technicolor’s imbibition prints, they were less costly to produce and generally more economical in small print runs.
Unfortunately, the introduction of Eastmancolor in the Fifties laid the foundation for the problems in color fading we are seeing today. Hollywood and the public both accepted Eastmancolor without really noticing a difference in color quality. Each studio developed its own special variant of the Eastmancolor process, and the crazy quilt of color names that burst forth in the mid-Fifties captured the satiric eye of Cole Porter in a number for the film Silk Stockings:
If Ava Gardner played Godiva riding on a mare,
The people wouldn’t pay a cent and they wouldn’t even care
Unless she had glorious Technicolor,
Or Cinecolor, or Warnercolor, or Metrocolor, or Eastmancolor,
Or Kodacolor, or any color . . .
And although both the public and Hollywood seemed intrigued with the possibilities of the new Eastman-based color processes, no one was thinking of its archival qualities. As color researchers Henry Wilhelm and Klaus B. Henricks write in their forthcoming Preservation of Contemporary Photographic Materials:
“In the early days of Eastmancolor – in the late 1950’s – release print dye fading was not generally of serious concern to the professional motion picture industry because prints of major features were generally physically deteriorated from scratching and abrasion during repeated projection and handling, and were usually discarded before dye fading became a serious problem …. The dark storage stability of Technicolor imbibition prints is far superior to release prints made on Eastman color print film, and this fact alone would justify the added cost of imbibition prints for many applications …. Faced with increasing competition from Eastmancolor, Technicolor considered advertising its imbibition process on the basis of its excellent stability. However, permanence has never been a primary requirement of the motion picture industry, and Technicolor decided not to make a major issue of the permanence question . . . .”
The honeymoon between Hollywood and the new Eastmancolor processes was predictably short-lived. In less than a decade, evidence of bad print and negative fading would cause some film companies to invest in expensive “separations” of their major films to insure their longevity. Other companies apprehensively surveyed their libraries and found the color in some films too faded to halt.
Today, when all that remain of the great Hollywood studios are corporate skeletons and film libraries, the importance of color fading to a distributor who has active interests in television sales and theatrical reissues is crucial. Usually the urgency of a studio’s color fading problem is in direct proportion to the care they initially took in developing, printing, and preserving their films.
Perhaps the two studios best prepared to confront the color fading problem are Walt Disney Productions and Metro-Goldwyn-Mayer.
The Disney studio has kept expensive separations on all of its major releases, both animated and live-action. Disney also, as policy, keeps an original Technicolor imbibition print on each title as a check if a question of color balance or tone arises during a reissue. Ironically, as far as the animated films go, the Disney studio could actually consult the original animation “cels” if needed. Naturally, the Disney organization’s concern with preservation is well-founded, as Disney films have the potential to keep earning money on a reissue basis almost in perpetuity.
MGM has had a long history of film conservation, which is now paying off with remarkable results. Wes Meyers, who heads the MGM Film Library in Culver City, California, confirms that MGM originally made protection elements on all of their films. In a continuing program of conservation, the studio has converted over 800 features from nitrate to safety stock, and has also finished converting over 700 shorts. The Company is currently working on trailers, travelogues, and cartoons. MGM keeps its original camera negatives in its vaults in California. And to assure that no negatives are ever destroyed by a man-made or natural disaster, the company keeps duplicate elements in an abandoned section of a salt mine in Kansas.
Myers credits the success of That’s Entertainment, the popular compilation of highlights from MGM musicals, with giving the company’s conversion program a shot in the arm by proving the public’s unflagging interest in MGM’s archival material. In fact, MGM has, in recent years, done very well financially with theatrical reissues of its films, organized into retrospectives that feature a wide variety of titles from the MGM library.
But even a preservation program as comprehensive as MGM’s can fall victim to the ravages of color fading solely due to the passage of time. During an MGM retrospective at The Museum of Modern Art in New York City, a print of George Cukor’s stylish Les Girls (1957) was a ghastly vision in beige-on-beige. A recent tribute to Vincente Minnelli featured a print of Tea and Sympathy (1956) that looked as if it had been brewed right in the pot alongside Deborah Kerr’s solicitude. Both prints had a history of minimal use and careful storage (MGM even labeled Tea and Sympathy an “archive print”), yet they showed as much color damage as prints subjected to callous treatment.
A relative newcomer to the area of repertory cinema distribution is 20th Century-Fox. The studio has recently tested the financial waters of the re-release market by assembling a program of about forty features from its library and playing them successfully in San Francisco and New York City. Unfortunately, Fox has many problems with negative and print fading. Much of their trouble stems from the inferior processing DeLuxe Laboratories (now DeLuxe General), a wholly-owned corporate subsidiary, did on Fox releases of the Fifties and Sixties. Many archivists accuse DeLuxe of processing Fox films too quickly and sloppily, and insist that improper “washing” of the developing chemicals from the prints and original negatives, as well as mediocre quality control, have contributed to Fox’s current preservation problems. One executive connected with the Fox reissue program referred to DeLuxe as “the cheapest and worst lab in the business. They simply do bad work.”
Sid Samuels, an affable, knowledgeable man who is in charge of print control for Fox, admits that the company has had trouble making up prints on some of their older films. “It’s an ongoing battle,” Samuels said recently. “We’re fighting time and chemistry. Fox does not want to see any of its films disappear this way, but we’ve had to depend heavily on the trickeries of modern technology to recover some of our pictures.” Certain Fox films on the reissue program recently shown at Manhattan’s Regency Theatre, such as The Seven Year Itch (1955) and Centennial Summer (1946), appeared to be in relatively good condition. But several other titles show painfully obvious problems in negative fading and improper laboratory timing. Sections of The Virgin Queen (1955) and Carousel (1956) appeared yellowish, and the lighting of indoor and outdoor sequences in both films was off. Bus Stop (1956) had a yellow tone from beginning to end, as did sections of How To Marry a Millionaire (1953). Forever Amber (1947) showed distinctly bad tonal shifts and contrast imbalances throughout the film.
Many of these problems are directly related to the storage and treatment of the original negatives. Samuels confirmed that many Fox negatives from the Fifties are in poor condition, mainly because they were used directly for striking prints, rather than utilizing a duplicating negative – the customary process to avoid damage to the original. Samuels said that negatives from the same period can vary dramatically in quality: “One negative may look gorgeous. Another, from the same storage area and the same time period, throws out a print that is an abomination.” Obviously, negative wear and fading is a complicated process, and proper temperature and humidity in storage, as well as clean chemical handling of the original negative, are important keys to longevity.
Despite the considerable trouble Fox has had preparing prints on older titles for theatrical engagements, the studio was able to improve the quality of problematic negatives and prints when preparing material for television broadcast. That’s Hollywood, a current half-hour television series syndicated by 20th Century-Fox Television, makes liberal use of excerpts from Fox films of the Fifties, many of which are badly faded. Through the use of computer circuits, technicians at Fox television were able to control each of the primary colors on the clips separately and, in effect, remix the color electronically to compensate for any dye loss in the original material. Although this technique proved reasonably successful for television broadcast, it cannot correct problems in theatrical release prints.
While distributors are understandably concerned over the significant problem of color fading, their concern tends to overshadow the additional problem of color consistency that occurs when films photographed in one color process are printed in another.
When new prints are needed on films originally shot in three-strip Technicolor, it is common practice for distributors to make up an Eastman internegative and strike prints on Eastman film. In the Fifties and Sixties this was done mainly for economic reasons, but now that Technicolor has phased out imbibition printing Hollywood no longer has a choice. This change from Technicolor to Eastmancolor results in a dramatic shift in color tone and balance. Hues that were rich and “warm” in the original Technicolor will become brighter and colder, almost neon-like, in Eastmancolor. Serious shifts in yellow tones occur, and Eastmancolor is simply not capable of reproducing the distinctive, saturated primary colors that distinguished three-strip Technicolor. Eastman prints of The Red Shoes (1948) struck recently lack the deep blacks and reds of the Technicolor imbibition prints. New prints of Meet Me In St. Louis (1944) and On The Town (1949) lack the vibrant quality they originally had.
A celebrated example of the incompatibility of different color processes involved Luchino Visconti’s The Leopard (1963). Visconti’s well-known penchant for detail and insistence on authenticity prompted him to lavish extraordinary attention on the costumes and sets for his epic. The Leopard was originally processed by Technicolor in Italy, but when Fox distributed the film in Great Britain and the United States they struck prints in DeLuxe color from a duplicate negative. This seriously affected the subtle, tapestry-like shadings Visconti had striven for, and sacrificed values in color fidelity and definition as well. The indignant director promptly fired off a letter to The Times of London, complaining that the film was “processed as if it were a bright piece of Hollywoodiana.”
Whether the problem involves color consistency or color fading, there is little a director can do about it after a film opens. Francois Truffaut admitted, somewhat fatalistically, that the best he could do to insure good color quality in Farenheit 451 would be to check personally the prints scheduled for major first-run engagements in Europe and the United States. After that, he confessed, the matter is beyond a director’s control.
Cinematographers are equally exasperated by the proliferation of faded prints that continue to circulate for years. Gordon Willis, who has worked with several directors known as meticulous color stylists (Francis Ford Coppola, Alan Pakula, Woody Allen), was understandably volatile when asked about the lax preservation of prints and negatives. “It has not been traditional for studio executives to preserve anything but their own jobs,” he remarked. “But then again, the basic overall structure of the motion picture business is not quality-oriented. Studios and labs usually think of film in terms of yardage.” When asked if he had personally seen faded prints of his work after its initial release, he quickly replied, “I generally don’t watch a movie after I’m through with it. Perhaps during the first six months of an engagement I’ll sometimes pop into a theater to check the print, the projection, and the audience reaction. But I certainly have no illusions about what a print will look like five years after its release.”
Retired cinematographer Arthur E. Arling, who began his motion picture career in 1927, served as operative cameraman on Gone With the Wind and won an Oscar in 1946 for his work on The Yearling, echoed Willis’s comments on the powerless position of a cinematographer regarding print quality during distribution. “Cinematographers usually lose control after the answer print,” Arling commented during a telephone interview. “Once the answer print goes to the lab, that’s it. I sometimes take a look at my films when they turn up on television, and I can tell they have deteriorated.” Arling also observed that the passing of the studio system may have something to do with the problems in color consistency and fading that are cropping up now. “When the major studios existed, we had more time to work on perfecting the color in films we photographed. Don’t get me wrong – we were still under pressure. But there was a continuity in laboratory procedures and systems, and we would also usually get the time to do things our way. That era, however, is gone for good, I’m afraid.”
As directors and cinematographers voice concern, even despair, over the color fading problem, the world’s film archivists struggle against time to save as many movies as they can. Listening to archivists provides a lesson in verbal intonation; their conversation is punctuated with pregnant pauses and laced with flashes of facetious wit (including a mordant reference to Vincente Minnelli’s “Rust” For Life). But the detrimental effects of color fading on various film collections are spoken of without levity.
Only a small percentage of The Museum of Modern Art’s 8,000-film collection is in color. But Eileen Bowser, a film curator at MOMA, cites color fading as the biggest problem in film preservation today – she fully expects the problem to get worse as the years go by. The Museum is currently putting its films into cold storage at just below freezing, in the hope that an economic method of dealing with color preservation will be developed in the next few years.
The Museum is also involved in making new prints of all the David O. Selznick color films, including Duel in the Sun (1946) and The Garden of Allah (1936), from Selznick’s original three-strip Technicolor negatives. The new prints will be prepared under the supervision of film technician Ralph Sargent and his staff at Film Technology Laboratory in California, specialists in the painstaking art of archival print manufacture. Sargent, who previously prepared two-color prints of Paramount’s 1929 film Redskin, is familiar with the difficulties involved in film restoration of this type. “The essence of this business is not only technical finesse, but also judgment,” he commented. “Many of these films require frame-by-frame checking.” This is taxing work that the larger laboratories are not equipped for, and so archivists rely more and more on the few specialists in this field.
Larry Carr, Administrator of Film Preservation for the American Film Institute in Washington, is convinced that most of the color problems coming to light today are due to “cutting corners by processing the film too quickly. The chemical process is very complicated, and rushing it can be a disaster. Actually, all handling, storage, and use of a film contributes to fading.”
One of the world’s largest film collections is in the Library of Congress, also in Washington. David L. Parker, Technical Officer for the Library, cites the years 1955 to 1958 as a “disastrous period in color fading,” noting that their prints of Carousel (1956) and The Long, Hot Summer (1958) have turned “monochrome fuchsia with some poisonous green overtones.” He sympathizes, however, with the problems studios have in getting good laboratory prints on their back titles. “They will often want 400 prints of 400 different films, instead of 400 prints of one film. The bulk of all their back titles could be considered non-commercial, except for the occasional reissue or booking into a repertory cinema. The labs and studios don’t really have the time to deal with the problem of striking quality prints on this basis. It goes against the entire grain of the film industry, which is to get the most money out of a picture at the time of initial release. The studios are also loath to throw good money after bad in the case of a film that was a financial disappointment the first time around.”
Archives that store Hollywood films are not the only ones threatened by color fading. Even the National Aeronautic and Space Administration, which owns the most elaborate archive in the United States for the storing of original film and slides from the manned space flights, will eventually have to face the problem of color fading. “It’s a fact of chemistry,” lamented Richard Thompson, Chief of NASA’s Photographic Technology Division. “The public is certainly not aware of the situation, but we anticipate color fading will be a big problem. The bulk of our most spectacular space footage is in color.” He seemed unfazed when told that many film aficionados would rather throw all the space footage out and replace it with rare prints of Orchestra Wives.
Color fading problems are, of course, not confined to archives in the United States. Harold Brown, Chief Preservation Officer for the British Film Institute, reported that early examples of hand-colored or stenciled films have survived much better than color films from the Fifties and Sixties. In the future, BFI will have to rely mainly on cold storage and expensive “separations” of their films to prolong their color life.
In June 1979, The New York Times reported that the Swedish Film Institute had asked the Swedish Government for $1.6 million to be used for film preservation. The Institute is alarmed, The Times reported, that “even modern films are not safe from the ravages of time. New films by Swedish directors such as Ingmar Bergman, Jan Troell, and Bo Widerberg start to fade, almost imperceptibly to the layman, as soon as they are made, and no one really knows how to halt the process, which accelerates with the passage of time.”
Without exception, each archive questioned said that Technicolor imbibition prints were in almost as good a condition as when they were first struck, with no measurable color fading. They were all distressed that Technicolor had discontinued a print process that had provided archives with a reprieve from color deterioration.
Technicolor had continued to offer its superior imbibition process in the United States until 1975, and Hollywood opted for it on such large-quantity print runs as My Fair Lady (1964), Dr. Zhivago (1965), and The Godfather (1972 and 1974). Faced with waning interest and increased costs, Technicolor finally closed its Hollywood inhibition plant in February, 1975. Their Rome operations were discontinued on June 1, 1978, and the last imbibition plant, in London, was shut down on June 14, 1978.
The imbibition process did not go out, however, without a fitting hail and farewell. Some years ago, Eric Spilker, a film writer, historian, and sometime entrepreneur, was interested in re-releasing The Gang’s All Here, the flamboyant Fox musical featuring some of Busby Berkeley’s most hallucinogenic musical numbers. Determined to present the film to the public as it had originally been seen, Spilker arranged to have Technicolor make imbibition prints of The Gang’s All Here and, in 1973, opened it in New York. Critics were enthusiastic about the film – and ecstatic about the color quality. Pauline Kael rejoiced over “the electric reds and greens of 20th Century-Fox’s Technicolor”; Rex Reed stated that “this one has been preserved in a print so beautiful and richly endowed with fadeproof Technicolor, it looks like it has been kept in a drawer with Darryl Zanuck’s old socks.” This was a unique instance where the superior imbibition prints of Technicolor garnered a large share of attention and praise – solely due to the instincts of a film historian convinced of the very real assets of the process.
An amusing and perhaps ironic conclusion to the Technicolor imbibition saga concerns the fact that in 1974, the People’s Republic of China contracted with Technicolor, Ltd. of England to have a complete imbibition plant built in China. Operation of the plant began in 1977. The Chinese are using the process to make large numbers of release prints of documentaries and training films. This makes China the last place where Technicolor imbibition printing can be done. Recently, the Chinese let it be known that they would be delighted to contract work from U.S. archives or studios in need of imbibition prints. And so, as the Technicolor sun rises in the mystic East, we ponder the ultimate implications of a laboratory populated by hard-working, happy Chinese technicians, engaged in striking new prints of There’s No Business Like Show Business.
Alas, color fading is not a problem that engages the concern of the general public. What little attention the problem does receive is due in part to the efforts of Henry Wilhelm, a specialist in the history of color in film. Wilhelm lays much of the blame for color fading at the doorstep of Eastman Kodak, one of the world’s largest manufacturers of film and film supplies. “Ultimately, the responsibility certainly rests with Kodak,” he recently said. “Their dye stability could be improved at a relatively small increase in cost. They have also been negligent in informing the public of just what the projected life of their dyes is. But color fading is a touchy subject with them.”
Indeed, for many years Eastman Kodak avoided dealing with the problem by issuing blanket statements on the quality of their products and affixing a famous disclaimer to every can of film they sold, warning that the film was not insured against any changes in dye color.
Recently, though, Eastman Kodak has adopted a more progressive, informative role on the subject of color fading, and the company’s research has also been responsible for two significant new methods of dealing with the problem. The first involves new film stocks with improved dye stability, which Kodak estimates will last up to twenty times longer than their other stocks. Company spokesman Michael More said that the new stocks are currently available in 16mm, but that Eastman Kodak will make them available in 35mm upon request.
The second development at Eastman Kodak is a method by which faded transparencies can actually be restored to their original color. This complex process is akin, at least in rudimentary theory, to the electronic restoration used by Fox to color-correct faded film clips. The three primary colors are separated and re-combined in their proper ratio, restoring the transparency to its correct color. It’s possible that, in the near future, a variation of this process will be developed to help restore film negatives that have been badly damaged by fading.
In other segments of the photography world, there has been great excitement over a new system of printing color transparencies called Cibachrome. Marketed by the Ciba-Geigy Company through its photographic subsidiary, Ilford, Inc., Cibachrome has impressed photographers with its ability to render startlingly sharp and vibrant colors. The visual excellence of Cibachrome is even more attractive because it is between three and ten times more stable than Technicolor imbibition printing. Officials at Ilford, Inc. admit that Cibachrome is basically a system for displaying transparencies and prints under constant light, but say that it could also be used by studios and archives as a duplicating stock for film prints. If it were to be used this way, it would probably be the most stable method of storage as yet discovered. Although the company has no immediate plans to work motion picture technology into Cibachrome’s development, it reports that researchers at the company’s scientific headquarters in Switzerland are interested in pursuing further motion-picture development of Cibachrome.
A permanent solution to the problem of color stability could involve the use of holograms, and important research, partially funded by the National Endowment for the Arts, is being conducted in this area. The holographic process, when refined, would utilize lasers to record and reconstruct correct film color.
Whatever method proves to be the key to color preservation in the future, it is a simple reality that the international film industry cannot be expected to be the curators of its own past. The money machine will not be responsible for the dream factory. Henry Wilhelm adds: “Why should Hollywood care about color fading? The business of Hollywood is to make films. Asking them to make and preserve them may be giving them too much responsibility. I think an outside group, backed by the government, should keep good prints and original negatives for the studios, and allow the studios retrieval rights to their films.”
The idea of a national repository is an attractive one, and could probably be funded for surprisingly little cost. Wilhelm suggests that a slight increase in copyright fees could cover the entire tab. But in an age when distributors are becoming sensitive over issues of film piracy, it’s rather doubtful that they would hand over negatives and prints to any national clearing house. Still, the studios have good relations with the archives: MGM gives its original three-strip negatives to Eastman House; Fox contributes prints to the U.C.L.A. Film Archive and The Museum of Modern Art; Paramount and Universal deposit hundreds of titles in the Library of Congress.
There is also the possibility for corporate action. Recently, the Philip Morris Company began an ambitious program to distribute 100 films throughout the United States as part of a promotional campaign for their Benson & Hedges 100s. One title, Hitchcock’s To Catch A Thief, was found to have an original negative in very bad condition. Philip Morris not only picked up the tab for all the restoration on the original negative, but paid for a duplicate negative to be deposited with the American Film Institute. It’s this type of forward-thinking corporate involvement that could lead to companies “adopting” films and restoring them, while getting favorable media attention for their efforts. IBM, currently sponsoring a television series entitled Movies To Remember, should realize that there won’t be any movies to remember unless steps are taken now to insure their survival in the future.
By whatever means the problem of color fading is ultimately solved, it is obvious that a problem of such magnitude will require the hard work – and hard cash – of distributors, archivists, laboratories, and researchers. Color fading is a problem that was created by more than one segment of the film industry, and it is a problem whose solution will be found in the combined efforts of all groups involved. One thing is clear: the task is formidable and the deadline is now.”
(O’Connell, Bill (1979): Fade Out. In: Film Comment, 15,5, pp. 11-18.)
“To the Editor:
I read Bill O’Connell’s excellent article in Film Comment, “Fade Out” (September-October), with a great deal of interest, since color preservation has always been a concern of mine. As you well know, the urgency and importance of this issue cannot be underestimated. Color preservation affects everyone who seriously concerns themselves with film. From filmmaker to cineaste, preservation of color film must become a public issue, to be dealt with in a serious, productive manner.
I would like to capitalize on your article and, hopefully, form an alliance between the Director’s Guild of America, the Academy of Motion Picture Arts and Sciences and the American Film Institute to fund research for alternative methods of color film processing, as well as an all out effort to save color negatives from further deterioration. I have just returned from Japan, where new film techniques are being actively developed and researched. The Japanese laser technology is astounding, and its implications for future film production are truly remarkable. The possibilities that exist for color film preservation are, without doubt, plentiful, and we must take some form of affirmative action toward solving the color problem. Laser technology and holograms are here, and the future is now!
I cannot understand an industry that promotes new directors and new films without regard for the built-in obsolescence of those new movies. Although I am well aware of the expense involved in the present method of color preservation (three black-and-white negatives). I do not believe that a less expensive method is unavailable. How can we sit back and allow a classic film, 2001: A Space Odyssey, to fade to magenta? Far worse, can we continue to live with the problem and design our films around it as George Lucas, by his own admission, has done with Star Wars, a film color-designed to appear faded.
As a director and a filmgoer, I firmly believe this issue must be forced to resolution. My own work has been severely affected, in that New York, New York was made to look like a Technicolor imbibition film. Within five years, its color will have faded beyond any recognition of the original concept, and the film will suffer for that loss. My present film, Raging Bull, was shot in black-and-white in order to avoid the color problem entirely.
Finally, I believe that directors, film students, and the Academy must form a unified front to combat the problem. Through benefits, fund-raising, publicity, demonstration of the problem, and if need be, militant action, we must band together to face the issue and solve the problem. I personally offer my services, time, and finances to this cause, in an effort to motivate my colleagues and friends to action.
If your readers have any suggestions that will support my efforts, I would like to hear them. I thank you for your contribution in bringing this problem to the public attention.
(Scorsese, Martin (1980): Letter. In: Film Comment, 16,1, 1980, p. 79.)
“During the many years three-color subtractive reproduction has been in use, many cases have arisen in which the results obtained by straightforward means were not satisfactory and methods for improving these results have been sought. The technique of masking which was invented by Albert1 in 1899 has become one of the important means of effecting improved color reproduction. A great deal of work has been done in developing practical techniques of masking to give better results; and the possible application of masking to the establishment of conditions for exact color reproduction has been investigated. In fact, masking has become an integral part of most of the theories of exact three-color reproduction.
A more recent method of improving color reproduction is the use of “colored couplers.” These materials have been described by Hanson and Vittum,2 and by Vittum, Sawdey, Herdle, and Scholl.3 The usefulness of such couplers is limited to films, such as Kodacolor or Ansco Color, which contain couplers that react with oxidized developer to form dyes. If the coupler which is originally present in a given layer of such a film is colored, having absorption in the desired region of the spectrum, and if during the coupling or dye-forming reaction the color of the coupler which is converted to dye is destroyed, the final color-developed image will be composed of a negative (or positive, if reversal processing is used) composed of the coupled dye image and a positive (or negative) composed of the residual unused coupler.
The application of this type of coupler to color photographic films requires a knowledge, of the optical characteristics of the colored-coupler images and the control of these characteristics so that they conform with the known practical requirements for producing color photographs.
1 E. Albert, German Patent 101,379 (1899); German Patent 116,538 (1900).
2 W. T. Hanson, Jr. and P. W. Vittum, J. Phot. Soc. Am. 13, 94 (1947).
3 Vittum, Sawdey, Herdle, and Scholl, J. Am. Chem. Soc. (to be published).”
(Hanson, Wesley T., Jr. (1950): Color Correction with Colored Couplers. In: Journal of the Optical Society of America, 40,3, 1950, pp. 166–171, on p. 166.)
“However, while colour was briefly highlighted in the 1950s, as part of cinema’s campaign to win back audiences from (black-and-white) television, developments such as Cinemascope would prove detrimental to Technicolor, whose dyes were not well suited to its processes. Moreover, Eastman-Kodak had by then developed a cheaper and less complicated process, and Technicolor gradually lost its dominance.5
5 Other successful three-colour systems on the market included the German Agfacolor, and its Russian derivative Sovcolor, the Italian Ferraniacolor, and the Belgian-French Gevacolor.”
(Everett, Wendy (2007): Mapping Colour. An Introduction to the Theories and Practices of Colour. In: Wendy Everett (ed.): Questions of Colour in Cinema. From Paintbrush to Pixel. Oxford: Peter Lang, pp. 7–38, on p. 22.)
“Not until the legal break-up of the Technicolor monopoly was demanded by the U.S. Department of Justice in 1947 and the invention by Eastman Kodak in 1949 of Eastmancolor, which – unlike Technicolor – did not need to be shot through a special three-process camera, could the number of colour, films released increase markedly. Between 1947 and 1954, the number of coloured films produced in Hollywood rose from 12 per cent of the total in 1947 to almost 50 per cent in 1954 (Cook, 462).
But then, repeating the pattern of expansion and subsequent contraction that had happened in the past, the percentage of films produced in colour actually declined once again.
Cook, David, A History of Narrative Film, New-York: WW. Norton, 3rd ed. 1996.”
(Stokes, Melvyn (2009): Colour in American Cinema. From The Great Train Robbery to Bonnie and Clyde. In: Raphaëlle Costa de Beauregard (ed.): Cinéma et couleur. Paris: M. Houdiard, pp. 184–192, on p. 187.)
“The cause, and the cures, for the problem of colour fading are inevitably bound up with economics. Until the 1940s, ‘colour’ in Western cinema almost invariably meant Technicolor. Prints of old Technicolor films are of course subject to the ills that befall all nitrate material; but the three-strip Technicolor system itself was a stable one, and although the colours will not last for ever they are not subject to any kind of rapid fading. In the early 1950s, the system changed. Bill O’Connell puts it succinctly in an article in Film Comment (September/October 1979), one of several informative articles in American magazines which predated the Scorsese campaign. ‘Eastman Kodak announced the development of a “multilayer” film which already contained the three dye-sensitive layers necessary for colour. This not only did away with the need for the bulky three-strip cameras, but also changed Hollywood’s method of striking prints. Until then, Technicolor struck prints utilising the “imbibition” process, in which the three different colour dyes were each applied separately to the film base… But with Eastman’s new “multi-layer” film, colours could be obtained quicker and easier. Although the Eastman colour prints were inferior in quality to Technicolor’s imbibition prints, they were less costly to produce and generally more economical in small print runs.’
The chemistry of colour film is by no means simple for the layman to follow; there are many factors involved, and of course there have been changes and advances in the last quarter of a century. But the basic problem of the instability of the now almost universal multi-layer system remains. (Eastman Kodak now has a somewhat more stable print stock. It would add several cents a foot to the cost of a print and is consequently not in use.) Unofficial information from Eastman Kodak, supplied to the Scorsese campaign, shows that movie colour stock has a shorter life expectancy than that used for photographs, and that within six years ‘a 10 per cent colour density loss affecting one or more of the principal dyes’ is predicted, though there is also a suggestion that a 10 per cent loss may not be too serious. But of course any imbalance in colour affects the whole, and if the blue and green fade the result will be an overall effect of magenta or pink (Visconti’s ‘pink leopard’). One of the more graphic quotations accompanying the Scorsese campaign stresses this: ‘After only five years, the blue is leaving the waters [of Jaws] while the blood spurting from Robert Shaw’s mouth gets redder and redder.’”
(Anonymous (1980): Colour Problem. In: Sight and Sound, 50, pp. 12–13, on pp. 12–13.)
“Hollywood relied on quick turnover of a large-scale production, and it would be years before TV screenings, cassettes and the whole opening up of the re-release market alerted the studios to the commercial value of the past. In the 50s, consequently, an impermanent art turned cheerfully to an impermanent format. Eastman Kodak leased out its process (as Warner Color, DeLuxe Color, etc.) and it became more and more general.”
(Anonymous (1980): Colour Problem. In: Sight and Sound, 50, pp. 12–13, on p. 13.)
“1981 has apparently been designated as the year of the colour film. If this had not already been decided, in whatever quarters these things are arranged, Martin Scorsese would probably have ensured it, with his characteristically vociferous and well-publicised campaign to draw attention to the fast fading colour on the modern screen. By Scorsese’s account, it’s virtually fading away before our eyes. ‘I have witnessed the deterioration and sometimes the destruction of most films I have seen. With the introduction of Eastman Kodak colour film in 1950, any hope for colour stability vanished. All films made in the Eastman colour process are about to deteriorate beyond repair. Some have already done so. Methods of restoration are so costly that if a film is not considered important it is left to die.’ And, in Scorsese’s terms, there are no such things as unimportant films: ‘All film must be saved. No committees should decide which film lives or dies, whether or not TV commercials are less important than movie trailers…’ Followers of the late, great Henri Langlois would agree.”
(Anonymous (1980): Colour Problem. In: Sight and Sound, 50, pp. 12–13, on p. 12.)
Deux ou trois choses que je sais d’elle (FRA 1966, Jean-Luc Godard)
“Jean-Luc Godard through his work is a stunning exception. Scanning his color films one immediately senses an unparalleled rigor in the organization of color. Overall, one might say, the color appears “artificial” or stylized with respect to the more familiar “natural” or postcard color of traditional films. But this generalization is unsatisfactory. I shall attempt to be more exact about the color strategies of Godard-, and the consequences of these strategies, by centering on one film, Deux ou trois choses que je sais d’elle [Two or Three Things I Know about Her] (1966-67).
I propose to analyze Godard’s palette in terms of four major tendencies: color tends to appear as a certain type of solid color, in a regular shape, with an arbitrary relationship to the surface of its object, and in primary opposition to other colors. It is important to note that not every color in Deux ou trois choses fits these specifications. This demonstrates that pertinent oppositions, when they appear, become structurally significant in terms of Godard’s overall color system.
Godard in Deux ou trois choses tends to limit himself to brilliant, solid colors. By “brilliant” and “solid” I have in mind a collection of attributes. First, the hue of the color appears uniform, relatively large in area, and autonomous – a solid – as opposed to the mixtures of hues found in designs, prints, or plaids. Secondly, the color appears both light in lightness and strong in saturation, hence “brilliant.”
There is one major variant of the solid hue in Deux ou trois choses: the striped hue. The reason that stripes are designated here as a variation of solids, rather than as a polar opposite to solids, is that Godard selects such a peculiar type of striped hue. Broad, evenly spaced stripes of one solid color, are alternated with similar stripes of a second solid color. The effect is of two brilliant solids juxtaposed equally edge to edge with neither one dominant and both distinctive. Indeed, striped patterns are employed in exercises by art students to minimize the influence of area and shape and to promote the effects of color.10 As we shall see, a principle of color equality is crucial to Godard’s color system.
These special stripes are seen in Deux ou trois choses for example, in the boutique when Juliette takes off her striped raincoat to hold up a striped shirt against a background of numerous shelves of striped clothes. And a red-and-white striped towel in a bathroom matches the red-and-white stripes of the American flag stenciled on John Bogus’s white T-shirt. The towel is also reflected in a mirror where we see a shirt striped in the complementary colors, yellow and blue-purple.
What are the consequences of using brilliant, solid colors? First, the consistent use of highly saturated colors is unnatural. Generally the colors of nature and everyday life are relatively unsaturated.11 This fact is exploited most by advertisers who favor highly saturated colors for their attention value in billboards and posters; indeed, advertising material itself is conspicuous in the films of Godard.
Secondly, with respect to lightness, Godard follows the so-called natural order of light values. The colors of the visible spectrum at full saturation are not all of equal lightness. Yellow is the lightest, blue-purple the darkest, and between them – arranged in descending semicircular sequences on a color circle – are the hues orange, red, red-purple, and purple; and the hues yellow-green, green, blue-green, and blue. Values of natural light and frequently the coloration of plants and animals follow this order. Consequently, there is a powerful expectation in the viewer that the light values of a composition will follow the order in which they appear in nature; a dark yellow (olive brown) paired with a blue (of the baby blue variety) seems “unnatural” or, similarly, an orange with pink or, as in the final bedroom scene of Deux ou trois choses, dark green with pink. As indicated, though, Godard generally respects the natural ordering of lightness in color, partly because he chooses highly saturated colors (each hue reaches its greatest saturation at its natural lightness); but, more importantly, because his method is to restrict the possibilities of color, selecting only a few elements to construct a disciplined color system.
Lastly, it is apparent that Godard rejects color schemes in which saturation and lightness mutually vary. One way these two variables may interact is through what Wilhelm Ostwald termed a “shadow series.” By this convention a single hue undergoes a series of precise changes which are meant to represent the color gradations of an object as it models from light into shadow.12 Thus the greens in a forest will progress from light and saturated near a light source (or apparent light source in the case of painting) to a darker, more grayish green in half-shadow to a green approaching black in deep shadow. The filmmaker, of course, unlike the painter, may actually use one or more light sources or filters in order to model colors in this way or to produce variations.
Godard does none of this. He typically uses high key, featureless lighting which suppresses shadow and creates flatter colors with less spatial or volume effect (chiaroscuro). Indoor scenes are as brightly lit as exterior daylight scenes. The traditional film, on the other hand, usually lights its objects to create depth, aligns its light sources with the mise-en-scène to emphasize “natural” sources (e.g., an open window, a lamp, fireplace, television set), and contrives its compositions so that the center of interest is accentuated by a sharp contrast with the surrounding light values. The uniformly lit objects of a Godard film, however, seem to glow on their own and are not so easily integrated into a natural, three-dimensional space. Also, the elimination of shadows intensifies and separates colors because the shadow from a colored object is not black, as is commonly thought, but approaches the complementary hue of the colored object. Eliminating these interfering shadows heightens the glowing flatness of surfaces, accentuates edge contrast, and imposes a certain equality on the elements of the composition.
So far I have discussed Godard’s use of lightness and saturation with respect to a uniform, solid hue. I now consider which hues are chosen and how they are ordered. This subject is usually examined by color theorists under the heading of color “harmony,” which is defined as a specific set of relations among the colors while all other relations create disharmony. The problem is that virtually every theorist has a different system of classifying colors and hence different criteria for the relations and harmonies among colors. The three variables or “dimensions” of color have been developed into numerous color systems which have taken the solid shapes of – among others – a rectangular chessboard, pyramid, double cone (Rood), toy top (Ostwald), sphere, irregular sphere (Munsell), and what resembles a smashed grapefruit (1SCC-NBS system). A “slice” of such a solid specifies basic hue relations. These slices have appeared as five, and six-pointed stars, “ring-stars” (Ostwald), pentagons, triangles,13 and, the most popular, circles, as well as combinations of these and others. To further complicate matters, there is no agreement on how many or which colors to include in a system.14
Systems are also constructed in terms of different primary colors. A primary color is one which may be combined in various amounts with other primaries to generate (almost) all the remaining colors of a system. The physical or light primaries (which mix toward white light) are red, blue, and green. The pigment or painter’s primaries (which mix down toward black) are red, blue, and yellow. The psychological primaries (which mix in vision toward gray) are red, blue, yellow, and green. It should be noted that there are still other primary systems, and that the above colors are not the same from system to system; for instance, the red physical primary is not the same as the red pigment primary.
It would seem that Godard chooses something akin to the painter’s primaries, selecting a brilliant, indivisible red, blue, and yellow. But, most startingly of all, a majority of the compositions in Deux ou trois choses are based on these three hues alone without the use of other, intermediate hues. Green rarely appears, although it is used decisively, for example, in the trees around the car wash/gas station, in the picture above the bed in the closing scene, and in the grass background of the final shot. Other colors are conspicuously absent. Thus, Godard tends to restrict color to brilliant and solid painter’s primaries.
One dominant color composition throughout the film involves the three colors red, white, and blue. First of all, this color combination has an inescapable cultural connotation. These are the national colors of the United States, Britain, and France. We see in the film the United States flag being worn by Johnny Bogus and an intercut of a poster with the letters U, S, and A colored red, white, and blue respectively. We also see the Union Jack painted on the lips of a model in magazine photograph. And of the opening thirteen title shots, all but two (which are green) are laid out in the order of blue, white, red. The letters of the end title (“Fin”) follow this order as does the French flag which contains these colors in three vertical blocks, left to right.
Other compositions employing these colors include the close-ups of Robert’s notebook (red ink on white paper ruled into blue squares); shots of the overpass (blue pavement, white curbs, and a red crane below); an extreme low angle of two derricks (one red, one white, against a clear blue sky); the frequent shots of a Mobil gas sign; and the appearance of various service stations. One of these latter shots in particular shows vast expanses of white concrete, blue gas pumps with red and white trim, an air pump in red and white, an attendant in blue uniform, red foreground flowers, two cars – one red, one white – and all the colors brilliant solids. There is also an extraordinary shot of Juliette lying in bed with, from left to right, a blue top, white sheet, and red blanket (cf. the colors of France). The next shot is of her son, composed in the same colors, but rearranged in the frame.
What are the consequences of color schemes based on the painter’s primaries red, blue, and yellow, and red and blue separated by white? First, there are no intermediate hues (nor intermediate values of light or saturation) to act as “transitions” between the primaries. Two colors are continuous in this sense when each contains something of the other. Thus, brown – which is, strictly speaking, not a hue but a dark shade of yellow, orange, red, or magenta – has the potential to link a number of colors. Transitional colors were important for the French Impressionist painters because they believed that, in scanning a painting, the eye follows the routes of least visual resistance; hence, patterns of transitional colors were called “passage.” Whether or not the theory is plausible with respect to eye movement, it does account for a system of differences and relations among the colors. According to this theory, then, the primary hues in Godard’s compositions are highly discontinuous and isolated. In fact, the juxtaposition of discontinuous colors serves to further increase the perceptual contrast of the two colors.15 Black, white, and gray – since they contain only the attribute of lightness – may mediate between any two colors but only in this dimension. The white of red, white, and blue combinations in Deux ou trois choses, however, is far too light with respect to the red and blue, and thus it serves to separate rather than bridge the two colors.
Though isolated, Godard’s colors are, nevertheless, balanced and symmetrical. The general rule of hue harmony states that hues should be either near one another or else far apart on a color circle. In this sense of “regularity” the concept of harmony is a requirement of form – a symmetry or repetition of units which assures that a work is a construction of meaning, not the result of chance.16 The traditional harmonies or forms are expressed in terms of degrees on a given color circle17 (cf. the 30-degree and 180-degree rules of spatial continuity in classical Hollywood films) or, most often, as various geometrical shapes rotated inside a color circle (straight lines, equilateral and isosceles triangles, squares, etc.) or spatial figures cut from a color solid. The primary colors of Godard might then be related as the points of an equilateral triangle – which neatly illustrates that the painter’s primaries, red, blue, and yellow, represent the strongest contrast of hue.18
These primary colors may be balanced, but they are not passive or static. This brings us to a second consequence of using primaries (the first was an isolation and discontinuity of color based on the absence of transitions): a tension is created between color areas. The tension is that of reds versus blues, the so-called warm/cool opposition.19 The contrast is sharpened by Godard’s tendency to limit color to brilliant solids. With primary colors, the tension is red and yellow versus blue; in red, white, and blue compositions the tension is at its most elemental, red versus blue.
The opposition of red and blue is often stated in spatial terms: reds advance toward the viewer, blues recede. This effect finds support in the well-known spatial cue of aerial perspective (the blueness of distance) and the fact that the lens of the eye is subject to chromatic aberration, which means that red light is focused behind the retina (as in far-sightedness) and blue light in front of the retina. Thus the tension engendered by red and blue appearing together begins as an optical stress in the musculature of the eye as the lens alternately tries to focus first one, then the other color. Two colors – green and magenta – focus directly on the retina and are spatially neutral; hence they are said to be “restful.” Godard’s system therefore creates isolated colors, equal and in opposition, with a selective use of green as mediation.
The warm/cool color dichotomy, established as difference, allows the text to construct meaning. The culture has already assembled a vast storehouse of possible (past) connotations (emotions) of the red/ blue opposition, such as, fire/ice, near/far, violent/tranquil, opaque/transparent, heavy/light, dry/wet, and more.
Repetition of color forces one to perceive it as a material element capable of entering into combinations and exchanges which coalesce into a system. Thus Godard’s color serves notice that nature will be radically reassembled, not imitated, and in a way which questions the dominant forms of assembly or representation. Color becomes a catalyst.
Regularity of shape
So far I have considered Godard’s color in terms of a selection of a limited number of solid hues of light value and strong saturation. I turn now to the shape of these colors and their placement within the frame. This is a complex subject and so 1 will sketch only some Godardian general tendencies.
The compositions of Deux ou trois choses tend to reveal strong horizontal and vertical lines and colors tend to appear in blocks and regular shapes, such as squares and rectangles. The preference for solid, uniform colors or stripes, noted earlier, contributes to the horizontal and vertical segmentation of the image. There is a shot, for example, of Juliette crossing the street prior to her entrance into the boutique. We see across the street six rectangular billboards. The first three contain no advertising, and each is painted in a brilliant, solid primary: red, yellow, and blue. The fourth is empty and unpainted. The last two are covered by shreds and tatters of various color posters. Many interior scenes contain posters which confine blocks of color against white, blank walls. In a 360-degree pan shot from the courtyard of a housing project, we see mostly the whites and grays of cement and skyline except for two large rectangular sections of wall that are painted a brilliant red. In the car wash, the car emerges a shiny red between two walls of bright yellow. And at another point in the film a woman wearing a sweater with horizontal blue stripes stands against a wall of three posters: one red, one yellow, and one blue. (Appearing on the blue poster are the words “color vision.”)
The repeated use of color in regular shapes tends to breakdown the “objective” image into a more formal, graphic segmentation. Natural contours tend to be replaced by a gridwork of lines that enforce not only the man-made (e.g., construction sites, cranes, buildings, advertisements, commercial signs, products) but also the abstract, the constructed – that which cannot be natural. The later paintings of Piet Mondrian, which are composed solely of rectangular color blocks, map the extreme of this strategy.
There is a major variation of these large color blocks throughout Deux ou trois choses. In the variation the same colors are used (brilliant, solid primaries) but are scattered into tiny pieces across the image. Again, the lack of intermediate colors promotes isolation and equality for the scattered hues. We see, for example, Juliette in the kitchen with fifteen or twenty household products – brightly colored boxes, bottles, and cans – scattered about. In the day care/brothel scene we see a table jammed with various brightly colored products. Later Robert appears against a massive advertising billboard which has been reduced to mere shreds and scraps of posters. In the cafe two men sit before stacks of multi-colored books including fiction, history, guidebooks, telephone books, and others in several languages, from which they randomly select and then quote a brief passage.
The last example illustrates that whatever breakdown has occurred in the visuals in terms of color is paralleled by a breakdown of the audio track and, perhaps, of narrative. The use of scattered color at various moments in the film seems to indicate a second disintegration of the objective image whereby the homogeneous color blocks have exploded into fragments. One is reminded of Juliette’s remark in the second scene when she says, “I feel as though I were being shattered into a million pieces when I dream…. When I wake up, I’m scared some of the bits will be missing.” In its episodic narrative and its color, Deux ou trois chose enacts a world, a social order, in the process of fragmentation.
Allied with the breakdown of color into equal elements is; a breakdown of the center of the image. The traditional film has often meant a tyranny of the center of the frame and central perspective, but in Deux ou trois choses there are compositions which reassert the equality of the sides of the frame. These effects are further emphasized by the expanded horizontal ratio of widescreen. We see, for instance, an extreme long shot of a construction site where a cement bucket suspended from a derrick moves at left frame, then off frame. The essentially static shot is now held until, finally, the cement bucket reappears briefly in the extreme top right corner of the frame (at a much greater distance from the camera). At the beauty parlour, Paulette’s soliloquy in close-up is punctuated by her hand breaking into the empty space at frame right. The rigorous spatial continuity of the second cafe scene depends principally on persons, reflected in a mirror, who appear in the top left or right corner of a shot. (Hence the continuity of this scene, and others in the film, is quite inaccessible to traditional reading procedures.) We also see from the distance of a long shot only Robert’s head at the lower left against a massive billboard. While the traditional film might shift a body off center frame, it would never show merely the head at that distance, but the entire body (cf. also the car wash scene).
These examples show how a tension is created between the centered and radically decentered compositions of Deux ou trois choses. The tension goes beyond a search for a new aesthetic balance; rather, it is in the end an attempt to reposition character with respect to the camera’s production of narrative space. Narrative no longer is the story of character, but is the story of a displacement of character. The political and ideological pressures and forces leading to displacement are the real subject matter of Deux ou trois choses. Prior to the first scene, an intertitle announces, “Eighteen Lessons about Industrial Society.”
What is important in the present context is that color plays an initial role in subverting the centered and hierarchical attention of a traditional reading. Colors may be positioned so as to compete with frame center. For example, a low angle shot of an overpass predominantly in whites and cement grays with a bit of light blue sky is marked by patches of brilliant red at lower left and at top right where a man, barely visible, paints a guard rail. Not only the position of color but other aspects such as movement, direction, shape, and size are important to the effect of color. Schemes have been developed which relate color and area into a new property called the “weight” or “force” of a color,20 but so far these schemes are rather crude. On the whole, there is scant theory of what happens as color is fashioned into more complex forms. What is evident is that Godard’s consistent use of regularly-shaped color blocks, with its dialectical opposite of splintered color, together with a stress away from center frame, promotes color to parallel and equal status with other elements. Systematic color then becomes a tool capable of contesting traditional stylistic and discursive organizations of a text.
If there exists a system of color in Deux ou trois choses, how does it relate to the narrative system of the film? What are the larger functions of color in the text? I begin with some functions that color does not perform.
It does not provide clues to, or mirror, the psychological states of the characters. For most critics this is the sole, or at least paramount, way color is interpreted in film. A weaker claim stemming from the same approach is that color sets the mood or tone of a scene – “vivid” colors for a “lively” scene, and so forth. The essence of this method is to shift adjectives and nouns from the narrative and attach them to the appearance of colors, as in “a chromatic sensuality that sharpened the hedonism of the plot” or “an aberrant grayness of alienation” that later “is modulated into a murky grayness to underline the barrenness of a mind drained of dreams and emotions.”21
All other uses of color for these critics are shelved in the broad category of “symbols” or “comments” by the director; that is, if colors are not emotional they must be “intellectual.” Whatever the categories may reveal about underlying assumptions of mind, they certainly reveal an impoverished notion of narrative – usually no notion except perhaps that of “plot.” There is no concept of non-narrative or counter-narrative elements in the text – that surplus of signification which recent analysis has demonstrated is so crucial to the work of the text.22 Hence, despite a claim that color may have other uses, most critics and textbook writers in their examples fall back to a psychological recuperation of the filmic text.
Color in Deux ou trois choses also does not function to enhance a reality effect. When Marianne (Anna Karina) and Ferdinand (Jean-Paul Belmondo) escape across the rooftops in Pierrot le fou (1965), we see a number of ventilating pipes painted in brilliant, solid colors: red, blue, yellow. The color of these pipes stands out (almost as if it is being quoted) because normally we expect exhaust pipes and smokestacks to be black or at least a dingy color. Unexpected or unfamiliar color tends to separate color from its natural object in the same way that unfamiliar size frustrates a monocular depth cue (size perspective). It is the gap between color and object that is important in Deux ou trois choses. The gap is created chiefly through the strategies we have already discussed: a tendency to limit color to brilliant, solid primaries of regular shapes. Moreover, as previously noted, the consequences of these strategies in almost every case were to contest a natural reading of the image.
For example, when Juliette and Marianne emerge from their car in the car wash scene (see color illustration), they are dressed in brilliant, solid primaries: Juliette in a simple blue top and yellow skirt, Marianne in a simple red top and green skirt (the single reference to green in the shot). These exact primaries are reiterated throughout the shot: red, by the car, various rectangular electrical switch boxes, trashcan, and pail; yellow, by various painted machinery; and blue, by a painted wall. The repetition of color suggests that color is not working solely in a system of natural reference; background objects are not lost in the background as mere decor, atmosphere, or “accents.” The split between color and object is evidence that a logic other than verisimilitude (the probable) is at work. Later, during the bathroom and sex scene with Johnny (a john), Juliette, walking back and forth, wears a blue travel bag over her head (she had just taken off her blue top) while Marianne, walking back and forth, wears a red travel bag over her head (she had just taken off her red top). These actions fill out a series of comparisons begun in the car wash scene, namely, a woman’s body is like a car’s body; and washing a car (labor) is like washing a body (in the bathroom) for blind labor (sex). The result is that commerce is seen as a form of travel, i.e. labor is a trafficking in bodies.
If color in Deux ou trois choses is nonpsychological, nondramatic, and nonverisimilar, what is it? In negative terms it is nonreferential, or self-referential; stated positively, it is an element of equal significance with other elements – a reversible field23 of permanences and permutations criss-crossing the text, capable of forming alliances at one point and contradictions at another. That is, a color is capable of connecting to various points in a text and helping to make patterns; it need not be confined to the surface of a specific object. This potential of color is seen most clearly in those moments when it is self-referential. In La Chinoise (1967) there is a close-up tracking shot past color samples from a paint chart. In Deux ou trois choses a giant commercial sign – Azur – appears several times and is painted blue. The word azur, of course, also names a shade of blue. (The color reference is enhanced by presenting the sign out of context.)
Color also achieves a certain independence when one object is seen to have more than one color. For instance, in a long held, extreme close-up of the hood of a car at the car wash we see green trees reflect as red, and blue sky as white. In the cafe the colors of a magazine advertisement are distorted through beer in a tall glass. In the hotel scene we watch as Juliette repeatedly turns a lamp on and off, changing the color of the lamp from white to blue and the wall from red to black; she says, “The image is permeated with meanings and memories.” These sorts of shots illustrate that in general Godard offers an image not as evidence for a fixed state of the world, but merely as a possible (cultural, personal, political, or common) description, a quotation or figure, which is limited and depends on further investigation. As Godard says in his whispered voice-over commentary in the cafe scene: “The frontiers, of my speech are the frontiers of my world…. [Whatever I say (or show in an image?) must impose limitations on the world, must make it finite [reduce it?].” (Cf. also Godard’s voice-over during the car wash scene and his inclusion of alternative camera takes within the plot of the film.)
Perhaps the most extraordinary use of color in Deux ou trois choses occurs in the final shot. The shot is, among other things, an elegant summary of the progress of color throughout the film. We see about thirty consumer products in boxes spread out on the grass (like the blocks of low-rent apartment buildings seen in the film). The colors of the products – essentially the primaries red, blue, and yellow – are set off against the green background of the grass and are modulated by a shadow series. As previously remarked, green is spatially neutral and “restful” while the shadow series contributes depth. Next, instead of a fade-out, Godard stops down the camera lens and holds on a darkness which completely loses the green, the shadow series, and the basic outlines of the products, but allows the scattered primary colors to continue to glow, almost independent of shape, referent, and nature.
At this point the breakdown of color through the film from nature to solid blocks to scattered pieces has become final. A parallel breakdown has occurred in the narrative where the traditional hierarchies of the causal-psychological chain – the “tight, economical”, plot – have been ruptured. The main character, for example, is first introduced as an actress, and the narrative veers promiscuously from drama to documentary to interview to soliloquy to guerrilla theater. But the breakdown in color and narrative can only be a starting point. As the voice-over commentary states at the end, “I’ve forgotten everything except that, as I’m going back to zero, I’ll have to use that as my point of departure.” We see that the color of Deux ou trois choses ultimately reveals a work in progress.
The psychophysics of color perception as well as the function of color in a textual system depend exclusively on context and relation. The aesthetic text no less than other texts is an instrument of social practice, and its meaning depends on a relationship with cultural conventions or codes. A set of color conventions that may be termed traditional or naturalistic may be employed as a background set in order to distinguish the color organization of Jean-Luc Godard in Deux ou trois choses, and the consequences of that organization with respect to traditional representation.
Godard tends to employ uniform hues, or stripes, which are light in brightness and strong in saturation (brilliant solids). Though his color generally respects the natural order of light values, the high saturations are more typical of advertisements than the low key colors of nature. Godard muses in whispered voice-over: “You might almost say that living in today’s world is rather like living in the middle of a big comic strip.” (Cf. also Godard’s ideas about the commodification of everyday life.) The shadow series – a special relation of lightness and saturation – is rejected in favor of high key, featureless lighting that suppresses natural depth and central perspective in order to create a flatter composition of equal elements. Also, the placement of off-center color blocks operates to subvert a centered and hierarchical attention.
The hues chosen by Godard are often the painter’s primaires – red, blue, yellow – which represent the extreme of three-hue contrast. Another frequent combination is the extreme of two-hue contrast: red and blue separated by white. This contrast is based on the warm/cool dichotomy and muscle stress in the eye. These extremes of contrast create colors in tension, though equal and balanced. The lack of intermediate colors (passage) accentuates the isolation and autonomy of color. The repetition of limited hue combinations acts to break down the natural image, which is expected to contain a multiplicity (a plenitude) of graded color.
The compositions of Deux ou trois choses tend to be strongly segmented by lines, often horizontal and vertical. As a result colors appear in blocks that act to destroy natural contours. A major variation is the appearance of the same colors but now splintered across the image as if color itself had broken down.
The colors of Deux ou trois choses cannot be read in terms of character psychology, the exigencies of drama, or of verisimilitude. Instead, color – divorced from its natural (probable) object through such strategies as the above – becomes a mobile element, and an element of equal significance with other elements of the text. The consistent use of color strategies means the construction of forms – positions and differences – which are the very foundation for the articulation of color in a pictorial system. Deux ou trois choses is one of the few color films which, to borrow Eisenstein’s phrase, is in color and not merely colored.
This, essay is a revised version of one that first appeared in Wide Angle, Vol. l, No. 3 (1976), pp. 20-31.
1 Jean-Luc Godard, Deux ou trois choses que je sais d’elle (Paris: Seuil, 1971); Jean-Luc Godard, Godard: Three Films (New York: Harper & Row, 1975). See esp. Alfred Guzzetti, Two or Three Things I Know about Her: Analysis of a Film by Godard (Cambridge, Mass.: Harvard University Press, 1981).
10 Josef Albers, Interaction of Color, Rev. ed. (New Haven, Conn.: Yale University Press, 1975), pp.48-50.
11 Color as Seen and Photographed, 2nd ed. (Rochester, N.Y.: Eastman Kodak Co., 1972), No. E-74, p. 54.
12 For the fine points of this convention and certain deviations which result in crowding of light values or exaggeration of saturation, see Arthur Pope, The Language of Drawing and Painting (New York: Russell & Russell, 1949), pp. 44-49, 99-106, 128-31.
13 For one modern version of triangular hue relations, see Rudolf Arnheim, Art and Visual Perception, Rev. ed. (Los Angeles: University of California Press, 1974), pp. 330-71
14 Most color circles are constructed with complementary colors – which neutralize each other when mixed – on opposite sides. Thus, there is a tacit assumption that colors seek an achromatic and absolute state, e.g., white light. Also, color circles include magenta (red-purple), which does not occur in natural light (it is mixed in the eye), and the circles invariably assume equal divisions for the colors (whereas reds and yellows comprise about 40% of the light spectrum).
15 Color has a temporal dimension. The perceptual mechanisms of simultaneous and successive color contrast – based on the formation of negative afterimages in the eyes – function to drive colors which adjoin or succeed one another further apart in the ways in which they differ. Camera and character movement, for instance, could be used to shift and juxtapose colors in simultaneous contrast while editing and optical effects, e.g., dissolves, would seem perfect tools to exploit the phenomenon of successive contrast. To my knowledge, no filmmaker has used these effects because negative afterimages are transitory and move with the eye and so produce an unstable, ambiguous and changing color.
16 Cf. Roland Barthes, “The Structuralist Activity” in Critical Essays (Evanston 111.: Northwestern University Press, 1972), pp: 217-18.
17 See Faber Birren, New Horizons in Color (New York: Reinhold Publishing Corp., 1955), pp. 32-33.
18 Johannes Itten, The Elements of Color (New York: Van Nostrand Reinhold Co., 1970), pp. 19-22, 29-31, 33, 72-74.
19 See William Charles Libby, Color and Structural Sense (Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1974), pp. 56-61, 67-68, plate 3.
20 See Itten, pp. 59-63.
21 Lewis Jacobs, “The Mobility of Color” in The Movies as Medium, ed. Lewis Jacobs (New York: Farrar, Straus & Giroux, 1970), pp. 191, 196. Godard’s color design may also be subjected to the standard approach, which seeks to narrativize all the colors. See, e.g., Paul J. Sharits, “Red, Blue, Godard,” Film Quarterly, Vol. 19, No. 4 (Summer 1966), pp. 24-9. What happens when colors in a Godard film fail to match plot or character? The critic then relies on irony as an explanation. “Godard, in his treatment of Camille’s garment hues [in Le Mépris, 1963 [Contempt)], seems to have broken from what he may have felt was a too-obvious color system. Camille, because she is in love with her husband at the beginning of the film, ‘should’ be wearing red; however, the cyclic motif that occurs in regard to Camille’s development has a particular irony that befits the irony of the film in general, particularly the ironic paralleling of Contempt‘s development with that of the Homeric Odyssey.” (p. 27) This approach to Godard’s color design has the ironic effect of making more conservative a director who is trying to create an alternative film practice.
22 Stephen Heath, “Film and System: Terms of Analysis,” Part I, Screen, Vol. 16, No. 1 (Spring 1975), pp. 7-77; Part II, Screen, Vol. 16, No. 2 (Summer 1975), pp. 91-113. See also Heath, Questions of Cinema (New York: Macmillan Press, 1981).
23 Cf. Roland Barthes, “The Third Meaning” from Image-Music-Text, trans, by Stephen Heath (New York: Hill and Wang, 1977).”
(Branigan, Edward (1976): The Articulation of Color in a Filmic System. Deux ou trois choses que je sais d’elle. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 170–182, on pp. 170–180.)
Pierrot le fou (FRA 1965, Jean-Luc Godard)
“When Marianne (Anna Karina) and Ferdinand (Jean-Paul Belmondo) escape across the rooftops in Pierrot le fou (1965), we see a number of ventilating pipes painted in brilliant, solid colors: red, blue, yellow. The color of these pipes stands out (almost as if it is being quoted) because normally we expect exhaust pipes and smokestacks to be black or at least a dingy color. Unexpected or unfamiliar color tends to separate color from its natural object in the same way that unfamiliar size frustrates a monocular depth cue (size perspective). It is the gap between color and object that is important in Deux ou trois choses. The gap is created chiefly through the strategies we have already discussed: a tendency to limit color to brilliant, solid primaries of regular shapes. Moreover, as previously noted, the consequences of these strategies in almost every case were to contest a natural reading of the image.”
(Branigan, Edward (1976): The Articulation of Color in a Filmic System. Deux ou trois choses que je sais d’elle. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 170–182, on p. 177.)
“Die Ausbreitung der Farbe
In den sechziger Jahren trugen weitere Farbfilmmaterialien wie Agfacolor, Eastman Color usw. zu einer größeren Verbreitung der Farbe bei. Nach der Einführung des Tons ist dies die zweite Revolution des Kinos. Die Farbe setzt sich zuerst in Western und Musicals durch, wo ihre Überlegenheit unbestreitbar ist. Anschließend erobert sie auch die Bereiche, die der Dramatik und Aussagekraft des Schwarz-Weiß-Kontrastes Vorbehalten schienen: den sozialkritischen Film, den Kriminalfilm und den psychologischen Film. Diese Revolution ist schleichend. Sie vollzieht sich nach und nach.”
(Borde, Raymond (1988): Die Filmarchive und der Farbfilm. Eine Einführung. In: Gert Koshofer: Color. Die Farben des Films. Berlin: Wissenschaftsverl. Volker Spiess, pp. 7–10, on p. 9.) (in German)
“In early 1950, Technicolor signed a consent decree that terminated its relations with Eastman. At least as important as the government suit was the fact that in 1945 the original Troland patents had expired. By the time that Technicolor had submitted to the suit, Eastman had already announced its own color negative process.23
It took about four years for Eastman Color to dethrone Technicolor: the three-strip camera was last used on Foxfire (1954). What gave Eastman Color the edge was that it could be used in any camera and processed and printed by generally conventional means. Within two years, most studios began to use Eastman Color to a degree, some claiming it as the basis of their ‘own’ system (e.g., Warnercolor, Columbia’s Super Cinecolor). In 1953, Eastman introduced an improved, faster negative stock and corresponding print and internegative stocks. At about the same time, studios discovered that Technicolor dye-imbibition printing did not yield enough resolution for the new widescreen processes. Thus Eastman Color was used to film The Robe, How to Marry a Millionaire, Beneath the Twelve-Mile Reef (all 1953), and other early anamorphic films. As of November 1955, most widescreen productions were shooting Eastman Color negative.24
Technicolor’s future was settled when Eastman entered the 35mm color market. A specialized firm concentrating on short-term and small-scale problems could not compete effectively with the basic-research program of Eastman Kodak. Eastman held thousands of patents, supported an immense laboratory, and invested millions in color research every year. (Eastman Color monopack grew directly out of the firm’s development of color couplers for still photography.) Most likely, Technicolor had long realized how precarious its control was and, expecting Eastman to devise monopack eventually, created the licensing agreements to give itself the first chance at the process. But color negative came too late, and three-color Technicolor gave way to a method which promised greater cost efficiency and compatibility with other innovations (e.g., widescreen).
23 The most detailed examination of Technicolor’s relation to Eastman Kodak is George E. Frost and S. Chesterfield Oppenheim, ‘A study of the professional color motion picture antitrust decrees and their effects,’ The Patent, Trademark and Copyright Journal of Research and Education, 4, no. 1 (Spring 1960): 1-39, and 4, no. 2 (Summer 1960): 108-49. See also Basten, Glorious Technicolor, p. 146; ‘Technicolor,’ Fortune, p. 54; Howard C. Brown, ‘Movies in color,’ IP, 8, no. 6 (July 1936): 26; Ed Gibbons, ‘Color,’ IP, 9, no. 6 (July 1937): 5-7; ‘Technicolor system,’ IP, 10, no. 1 (February 1938): 10; Ed Gibbons, ‘Color progress dominates 1939 technical horizon,’ IP, 10, no. 12 (January 1939): 9-10; ‘Technicolor answers anti-trust action,’ Technicolor News and Views, 10, no. 1 (January 1948): 1; ‘Color film suit settled in US consent decree,’ Los Angeles Times (25 November 1948): sec. 2, p. 2; ‘Eastman gets color patents,’ Hollywood Reporter (10 July 1951): 1, 4; ‘Government case against Technicolor terminated,’ Technicolor News and Views, 12, no. 1 (March 1950): 1-2.
24 ‘Color in the motion picture,’ pp. 164, 166; ‘Six companies now testing Eastman Color,’ Hollywood Citizen-News (10 November 1949): 19; Earl Theisen, ‘Notes on the history of color in motion pictures,’ IP, 8, no. 5 (June 1936): 8-9, 24; Don Hooper, ‘Negative-positive color,’ IP, 9, no. 8 (September 1937): 27-9; W.T. Hanson, ‘Color negative and color positive film for motion picture use,’ JSMPTE, 58, no. 8 (March 1952): 223-5; W.T. Hanson and W.I. Kisner, ‘Improved color films for color motion-picture production,’ JSMPTE, 61, no. 6 (December 1953): 670-2; Basten, Glorious Technicolor, pp. 149, 160; Frederick Foster, ‘Eastman negative-positive color films for motion pictures,’ AC, 34, no. 7 (July 1953): 322-33, 348; Robert A. Mitchell, ‘Color and its reproduction on film,’ IPro, 31, no. 2 (February 1956): 17; James Morris, ‘1954 seen as biggest year for color,’ IPro, 29, no. 1 (January 1954): 7-8; Robert A. Mitchell, ‘To which IP replies,’ IPro, 30, no. 8 (August 1955): 16; ‘1945 to 1955: ten years of progress in projection technology,’ IPro, 30, no. 12 (December 1955): 24, 38; ‘Summary of current widescreen systems of photography,’ AC, 36, no. 11 (November 1955): 676; ‘CinemaScope,’ International Sound Technician, 1, no. 2 (April 1953): 2; Robert A. Mitchell, ‘Anatomy of CinemaScope,’ IPro, 29, no. 6 (June 1954): 10; ‘Warner Brothers debuts “Warnercolor”‘ AC, 33, no. 3 (March 1952): 122; Edwin A. DuPar, ‘Warner-Color – newest of color film processes,’ AC, 33, no. 9 (September 1952): 384-5. See also R.M. Wiener, ‘Color film often doomed at birth – in the lab,’ Box Office (5 May 1980): 1, 5, 30.”
(Bordwell, David; Staiger, Janet; Thompson, Kristin (1985): The Classical Hollywood Cinema. Film Style and Mode of Production to 1960. London: Routledge, on p. 357.)
“Dank als Kriegsbeute beschlagnahmter Patente und mitgenommener Rezepturen sowie der Mitwirkung von Agfa-Fachleuten bei Konkurrenzunternehmen eroberte das in Deutschland ausgearbeitete Farb-Negativ/Positiv-Verfahren auch andere Länder: Aus Belgien kamen die sehr ähnlichen Gevacolor-Filme, aus Italien Ferraniacolor, aus der Schweiz Telcolor, in den USA folgten Ansco-Color und Kodak Eastman Color, in Japan Fujicolor.31
LITERATUR- UND QUELLENANGABEN
31 Gert Koshofer: Die Agfacolor Story, in: Weltwunder der Kinematographie, 5. Ausgabe, Potsdam 1999, S. 69ff.; Gert Koshofer: COLOR Die Farben des Films (1988), S. 109ff., 119ff.”
(Beyer, Friedemann; Koshofer, Gert; Krüger, Michael (2010): UFA in Farbe. Technik, Politik und Starkult zwischen 1936 und 1945. München: Collection Rolf Heyne, on p. 54.) (in German)
“Dave Davis was born in South Wales in 1921. After serving in the artillery in World War II he joined the Technicolor Labs in Denham in 1946 working in the administration side of the labs, taking care of customers’ requirements. He worked for Technicolor for thirty years, ultimately becoming production controller and then production supervisor. In this capacity Davis developed a system for organising the enormous volume of print jobs which the labs handled every day.
EXTRACTS PERTAINING TO COLOUR FROM BECTU INTERVIEW NO. 230
DATE OF INTERVIEW: 27 NOVEMBER 1991
INTERVIEWERS: ALAN LAWSON AND SYD WILSON
DAVE DAVIS: […] There was a competition between Eastmancolor and Technicolor dye transfer, because the facilities we were getting in the 1960s came sort of half and half. We got half of matrices and half of colour reversals, and the balance of the production schedule moved from being completely dye transfer to probably half Eastmancolor. Although we did all the scheduling, well with Eastmancolor, it wasn’t as complicated because you didn’t have backgrounds and things like that, and you couldn’t use overlays on printers, so all the things that made good films – the Rolls Royces of the industry – you could do. You had to photograph completely new bits of title into the film if you wanted to change it, and gradually with more people on the continent and over the world speaking English, more and more prints were shown in its [sic] English version.”
(Brown, Simon; Street, Sarah; Watkins, Liz (eds.) (2013): Interview. Dave Davis. In: British Colour Cinema. Practices and Theories. Hampshire: Palgrave Macmillan, pp. 123–130, on pp. 125–127.)
“Ab 1954 kehrt sich das Produktionsverhältnis von Schwarzweiß- und Farbfilmen in Hollywood um. Wurden 1951 erst 23,8% aller Hollywoodfilme in Farbe gedreht, so steigerte sich die Quote über 39,8% 1952 und 42,7% 1953 bis 1954 auf 58,4% (Waidekranz/Arpe 1956, 263). Dies hängt vor allem mit dem preiswerteren Mehrschichtenfarbfilm von Eastman Kodak zusammen, der dem teuren Technicolor-Verfahren jetzt schnell den Rang ablief.
Waidekranz, Rune/Arpe, Verner (1956) Knaurs Buch vom Film. München/Zürich: Knaur.”
(Penning, Lars (1988): Farbe im klassischen Piratenfilm. In: Karl-Dietmar Möller-Nass, Hasko Schneider and Hans J. Wulff (eds.): 1. Film- und Fernsehwissenschaftliches Kolloquium. Münster: MAkS, pp. 36–40, on p. 38.) (in German)
“WarnerColor – Newest of Color Film Process
Lab developed and field tested, new medium proves highly successful.
By EDWIN B. DuPAR, A.S.C.
EDITOR’S NOTE: WarnerColor is a negative-positive process, utilizing the new Eastman color films in which the natural colors of the scene photographed appear in their complementaries on the developed negative.
Prints can be made by two methods: either by contact on positive color stock or by making three color-separations on fine-grain panchromatic stock with proper filters, then printing from the separations in technique similar to lithography.
The speed factor of the new WarnerColor is 16, as against 12 for Kodachrome, making the former half a stop faster.
The development of WarnerColor has been pursued in two directions—in photography and in the film laboratory. In the former, Warner Brothers’ top directors of photography have lent their talents and knowledge of the photographic art. On the laboratory side, Fred Gage, A.S.C., Warner’s lab head, mothered the process to its present state of perfection. The process is exclusively Warner Brothers’. No plans for making it available to other studios have been announced.
WARNERCOLOR has become one of the most talked-about color film processes in the motion picture industry. At this writing, four pictures have been completed at Warner Brothers studio in the new WarnerColor, and we now can look back and evaluate the progress that has been made with this new and remarkable process.
I directed the photography on three of the initial four productions, and I sincerely believe that WarnerColor is the finest color film we have in the industry. It is certainly the most satisfactory that I have used in thirty years as a motion picture cameraman.
The colors are natural and true. Definition is extremely good, and extraordinary in shadows. Photography can be carried on in any weather. Extreme highlights do not bother the eye because glare is absent, even in intense sunlight or in a snow background. The film is unusually sensitive; it can be handled in the laboratory as easily as black and white film. No special equipment is necessary and rushes may be viewed the next day.
The tremendous progress the studio has made with the process is exemplified in “The Miracle of Our Lady Of Fatima,” the third WarnerColor picture now in general release. When I was assigned to direct the photography of “Miracle” I discussed the matter of makeup with director John Brahm and producer Bryan Foy, and it was determined that little or no makeup would be used on the players. Most of the cast represented peasants and children, and none wore makeup. Even when filming the crowd scenes, we asked the women to first remove their, street makeup – lipstick and rouge. This produced improved photographic results, because the true colors of the players’ features were brought to the screen.
A companion development in the WarnerColor process has been a new makeup for use with the film. Gordon Bau, head of Warner’s makeup department now has a light cosmetic specially for WarnerColor. When work first began on the film process, Bau set up an experimental laboratory just to explore and test new makeups for WarnerColor. He found that makeup could be eliminated almost entirely on players in all WarnerColor films, were it not for the day-to-day variations in players’ complexions. Some makeup, he found, is needed to equalize the skin color and conceal the slight daily changes that occur due to health, sunburn or windburn.
We discovered many interesting things about lighting with WarnerColor, too. “The Miracle of Our Lady of Fatima” lent itself particularly to low key lighting in the interiors, and I therefore kept most of the interiors at a low level. The early scene between Gilbert Roland and Jay Novello in the wine shop responded particularly well to this lighting treatment, being sharp throughout and with genuine natural color. Because of WarnerColor’s great depth of focus, most of our exteriors have a strong three-dimensional aspect.
The numerous scenes on the hilltop in which the image of the Virgin appears to the three children were done on the original negative in the camera; however, the Warner Brothers’ laboratory is now using the three separation negative process in printing these scenes.
In “Fatima” the emphasis was on characterizations rather than on action, which implemented the previous WarnerColor productions. Director Brahm, therefore, was painstaking in developing compelling camera studies – virtually Portraits – of the principal players, which may very well become the envy of contemporary still life painters, so successfully did we achieve new pictorial results with this color film process.
I am naturally delighted with the reception that has been given the Warner-Color photography of “Fatima.” Already I have received many letters and telephone calls about it – some from persons not connected in any way with photography. As it would be to any motion picture cameraman, it was a great satisfaction to me that so many people recognized the great difference and improvement in this new phase of screen photography.
Two days after finishing “Fatima,” I went on to new and even more interesting experiences with WarnerColor when I started shooting Warner’s “Springfield Rifle,” starring Gary Cooper. We started the production by shooting exteriors at the famous Lone Pine location site, filming on Mt. Whitney, about 9,000 feet elevation. From the cameraman’s point of view, this Warner-Color production was vastly different from any of the preceding three. We didn’t have any booster lights along because the story lent itself to many different moods, and the studio had decided to shoot in any and all weather – which we did. As an example of the shooting weather often encountered, one day we were filming in the snow in bright sunlight when the sky suddenly became overcast and snow started to fall. For the next three hours we had a full-sized blizzard, with snow reaching a depth of four inches. Director Andre De Toth took advantage of the situation for its pictorial possibilities and had us shoot one full sequence during the blizzard. It was so dark, I could scarcely get a meter reading – the meter needle barely moved – but we took a chance, with the result that we filmed scenes that never could have been obtained if staged under artificial storm conditions, indoors or out. This further testifies to the unusual qualities of Warner’s new color film process.
“Carson City,” starring Randolph Scott, was the studio’s initial production with the new WarnerColor process. Quite naturally, John Boyle, A.S.C., who directed the photography, ran into many problems impossible to anticipate, even though extensive laboratory experiments had preceded the initial test-in-production of the film. Those of us who had worked on the experiments were a little nervous at the start, for no matter what our confidence was in the new film, we also knew that the studio had a good many dollars invested in pioneering it. During the tests and experiments, there were times when we were straining to get an exposure, and still other times when we were not at all certain that our focal depth was correct. However, none of us at Warner Brothers was prepared for the surprise in finding every exposure perfect. We now realized that we had not even begun to tap the vast potential of the new film’s possibilities. As the dailies rolled in, we knew we had a winner.
WarnerColor has been in steady use now for little more than a year. The strides that have been made are tremendous; the improvement in color quality, contrast, etc., from film to film has been a revelation; but all this is all the more intriguing when you consider we still have a lot to learn about WarnerColor’s limitless possibilities. This film process, I am sure, will open up a whole new era to the motion picture industry – an era sure to see most all productions made in color at a cost scarcely more than that for black-and-white.
The WarnerColor production on which I am now working – “Back To Broadway,” starring Virginia Mayo and Steve Cochran – is entirely different from all the others, with most of the scenes interiors. The film is set up for a 450 foot candle key at f/2.8, but there is very great latitude either way.
I have been fortunate in having the same crew with me on each WarnerColor production including operator Lou Jennings and gaffer Vic Johnson. Their vast experience and splendid assistance have made it possible for me to devote more time to getting the most out of this new color film process.”
(Dupar, Edwin B. (1952): Warnercolor. Newest of Color Film Process. In: American Cinematographer, 33,9, pp. 384–385.)
Track of the Cat (USA 1954, William A. Wellman)
Dans le même temps où la couleur perd, par banalisation et fidélité croissante à la nature, quelque chose de sa distinction, certains cinéastes s’emploient à la réduire délibérément et comme arbitrairement. Une manière de nostalgie du bichrome ou des débuts du trichrome s’exprime là, qui bien sûr n’est pas un retour pur et simple aux schémas chromatiques primitifs, puisque cette nostalgie maîtrise les progrès techniques et les apports successifs de la couleur, mais avec la volonté, pour ainsi dire, de les tenir en bride. Il faut évoquer à cet égard Moby Dick de John Huston et Track of the Cat de William Wellman, qui fonctionnent de manière fort comparable.
Track of the Cat (1954), dont l’action est située dans le Colorado, déploie un paysage de western en CinemaScope, et des intérieurs primitifs, un peu de style shaker, d’où la couleur paraît presque absente. Sur la pellicule Technicolor, le noir et le blanc dominent: paysage de neige, glace, brouillard et brumes, et de sapins noirs; intérieurs où tout, les murs, les vêtements, les ustensiles, est blanc et noir, notamment une couverture qui servira de linceul. Le noir et le blanc s’étalent en larges taches comme sur une veste en peau de bovidé, ou constituent un motif monochrome à rayures genre prince-de-galles.
La règle semble ne tolérer qu’une exception, affirmée avec d’autant plus d’éclat: il s’agit du blouson rouge que porte Robert Mitchum, d’un rouge “indien” très vif comme on le voit souvent dans les westerns, avec une bande noire au milieu qui l’intègre en quelque sorte au système chromatique ambiant. Ce blouson servira aussi de linceul, et disparaît ensuite de l’action.
D’autres exceptions sont moins apparentes. En réalité, l’image n’est pas toujours, stricto sensu, monochrome. Des tons crème, la couleur du bois naturel, la carnation des visages composent une sorte de camaïeu. Les visages, les chevelures blondes ou châtain de certains personnages (l’un d’eux a, au contraire, une barbe noire et blanche), les yeux bleus… constituent autant de notes comparativement chaudes et colorées dans le système chromatique noir-et-blanc. L’intérieur d’un cercueil noir a un ton de bois naturel assez chaud, presque blond. Une bouteille de bourbon jette parfois, dans la salle des Bridges, une seule tache de couleur un peu vive et foncée. Vers la fin du film, quelques plans montrent le bleu du ciel, les conifères retrouvent une teinte plus verte que noire, la flamme jaune d’un brasier s’élève et équivaut plus ou moins à la tache rouge du blouson, dans la première partie. Renversement significatif: alors qu’au générique, un paysage blanc et noir était frappé de l’écusson W[arner] Brothers] en bleu et jaune vifs, la mention “the end” apparaît en noir sur un paysage encore monochrome mais où brille à droite la flamme jaune vif du brasier.
Le plus souvent, ces notations de couleur vive, qu’il s’agisse des éléments du “Pequod” ou des taches de sang, perdent bientôt leur éclat pour se ternir, se fondre dans le camaïeu de l’ensemble. A l’intérieur même du camaïeu, une évolution se dessine, de l’harmonie plus sombre (bistre/gris) de l’ouverture, où brillent par contraste les yeux extraordinairement bleus et lumineux de Richard Basehart, à la suite maritime, nécessairement plus claire, à dominante bleue/verte.
Cela montre, comme dans Track of the Cat, combien un tel parti chromatique est difficile à respecter. Le blouson rouge de Mitchum contraste, plus nettement que la flamme jaune, avec le paysage noir et blanc; il en va de même de l’œil bleu de Moby Dick, qui tranche moins nettement avec l’environnement maritime que les yeux de Basehart avec les intérieurs sombres du début. On en vient même à se demander si l’impression de monochrome n’est pas au moins en partie étayée sur des pilotis linguistiques, c’est-à-dire que certaines couleurs (le rouge, le jaune, le bleu…) sont faciles à nommer tandis que les nuances infinies du beige, du crème, de l’ocre, du bistre, des tons de chair, carnations et musculatures… n’évoquant pas une nomenclature précise, bien définie, sont ipso facto assimilées à un système plus ou moins monochrome, En ce sens, l’absence de couleur, dans Moby Dick comme dans Track of the Cat, serait plus ou moins l’absence de couleurs faciles à nommer.”
(Bourget, Jean-Loup (1995): Esthétiques du Technicolor. In: Jacques Aumont (ed.): La couleur au cinéma. Mailand: Mazzotta, pp. 110–119, on pp. 114–115.) (in French)
“D. 3.2.1. Farb-Negativ-Prozeß ECN-2
Vor einigen Jahren brachte Kodak neue Farb-Negativ-Materialien mit der Bezeichnung 5247 (35 mm) und 7247 (16 mm) heraus. Die Materialien sind im sogenannten ECN-2-Prozeß – einem Heiß-Prozeß mit kurzer Bearbeitungszeit – zu entwickeln. Die Bearbeitungsstufen im einzelnen gibt Tabelle Vll (S. 214–215) an. Bild 72 zeigt das Schema einer Entwicklungsmaschine für den Color-Negativ-Prozeß ECN-2.
D. 188.8.131.52. Verschleppungserscheinungen
Ein besonderes Problem ist die Verschleppung der dem Film anhaftenden Flüssigkeitsreste von einem Tank in den folgenden. Trifft man keine besonderen Maßnahmen, so wird etwa die doppelte bis dreifache Menge an Flüssigkeit vom Film mitgeschleppt, als von der Emulsion aufgenommen wurde. Während einerseits die Verschleppung von Wasser zu einer unkontrollierbaren Verdünnung der Bäder führt, führt die Verschleppung von Badflüssigkeit zu einer Anreicherung des Waschwassers mit Chemikalien, die eine unkontrollierte Nachreaktion oder eine Verunreinigung des Wassers zur Folge haben können. Um Fehler und Störungen dieser Art zu vermeiden, verwendet man Abstreifeinrichtungen beim Übergang des Films von einem Tank zum anderen. Sie wischen durch Abstreifblätter, Luft, Vakuum, Plüschrollen oder auch durch Absauger von beiden Seiten des Films die Flüssigkeit ab und verhindern somit den Übertritt in den folgenden Tank.
D. 184.108.40.206. Wässerungen
Eine große Ersparnis an Waschwasser kann durch eine “Kaskadenanordnung” der Wässerungstanks erreicht werden (Bild 73). Dabei fließt frisches Wasser in den letzten Tank der Entwicklungsmaschine, von dort in den jeweils davorliegenden bis zum ersten Tank der jeweiligen Wässerung. Da der Film gegen den Wasserstrom läuft, gelangt er von Tank zu Tank in saubereres Wasser.
D. 220.127.116.11. Trocknung
Die Trocknung des Films ist sorgfältig zu kontrollieren. Bei unzureichender Trocknung bleibt die Emulsion weich und klebrig. Bei zu starker Trocknung wird die Emulsion brüchig und kann abblättern. In beiden Fällen tritt eine Beschädigung der Oberfläche ein. Bei einwandfreier Trocknung ist der Film, wenn er die Hälfte des Trockenschrankes durchlaufen hat, bereits trocken, ohne klebrig zu sein. Vor dem Aufspulen muß er Gelegenheit haben, sich wieder auf Raumtemperatur abzukühlen. Nach der Abkühlung sollte das Material eine Feuchte haben, die mit Luft von 50% relativer Feuchtigkeit im Gleichgewicht ist. Ideale Trocknungsbedingungen erreicht man nur in Verbindung mit einer Vollklimaanlage.
D. 18.104.22.168. Umwälzung, Temperierung und Regenerierung
Um eine gleichmäßige Beschaffenheit der photographischen Bäder zu gewährleisten, muß eine ständige Umwälzung der Badflüssigkeit stattfinden. Hierzu verwendet man Pumpen aus korrosionsfesten Materialien, die die Badflüssigkeiten ständig in Umlauf halten (Bild 74).
Die Badflüssigkeit gelangt aus dem Bädertank (1) durch einen Überlauf (2) zunächst in ein Auffanggefäß (3). Von dort aus fördert die Umwälzpumpe (4) das Bad über ein Filter, das zur Reinigung der Flüssigkeit dient, über die Kühlung (6) und das Heizaggregat (7) zum Zulauf (8) des Tanks zurück.
Der Temperaturfühler (9) mißt ständig die Temperatur des Bades und gibt den Meßwert an die Temperaturkontrolle (11) weiter. Hier werden ständig die Ist-Werte mit den Temperatur-Soll-Werten (10) verglichen. Bei Abweichungen vom Soll-Wert wird Heizung (7) oder Kühlung (6) von der Temperatur-Kontrolle entsprechend gesteuert.
Die verbrauchten Substanzen der Bäder müssen ständig erneuert werden. Diesen Vorgang nennt man Regenerierung. Bei einer Reihe von Bädern geschieht das durch Zudosieren der verbrauchten Chemikalien aus einem Vorratsgefäß. Dabei ist zu unterscheiden, ob 35-mm-, 16-mm-Schichtfilm oder Blankfilm durch die Entwicklungsmaschine laufen. Dies stellt ein Meßfühler (17) bereits fest, wenn der Film in den Vorlaufspeicher der Maschine einläuft. Der Meßfühler gibt ein entsprechendes Signal an die Dosiersteuerung (12), die die Fördermenge der Dosierpumpe (13) bestimmt. Über die Dosierpumpe (13) gelangt das frische Regenerat (14) in den Bädertank.”
(Webers, Johannes; Westendorp, Kurt (1979): Einführung in die Kopierwerktechnik (XIII). In: Fernseh- und Kinotechnik, 33,6, pp. 213–215.) (in German)
Lust for Life (USA 1956, Vincente Minnelli)
“Failure to merge color with the elasticity and dramatic flow of the subject accounts in great part for the weakness of Moulin Rouge (1953) and Lust for Life (1956), whose color design stemmed from the palettes of Toulouse Lautrec and Vincent Van Gogh. Because the directors of these movies failed to realize that, on the screen, color can be structured in time and not just in space as in painting, they neglected to use color for more than its emblematic associations. […]
Yellows, oranges, reds, and black keyed the visual style of Lust for Life. But only toward the end of the film was the color pattern extended beyond individual shots to a succession that underlined the dramatic situation. As Van Gogh’s mind begins to falter and he becomes increasingly mad, the color progressively changes from shot to shot; the reds and browns appear less and less until there is only the pale yellow of a wheat field, which provokes an uncanny sense of foreboding and approaching death. When the artist finally dies, the yellow of the field is suddenly torn apart by the swift inundation of a flock of black crows. The impact of the abrupt contrast between yellow and black was striking and both dramatically and psychologically expressive. The scene gained an added overtone from the fact that the color scheme and composition which inspired it came from the artist’s own painting of a wheat field and crows, made at the very time of his own approaching madness and death.”
(Jacobs, Lewis (1970): The Mobility of Color. In: Lewis Jacobs (ed.): The Movies as Medium. New York: Farrar, Straus and Giroux, pp. 189–196, on p. 193.)
“Over the past twelve years, numerous color motion-picture productions have been made on Eastman Color Negative Film and many are now familiar with its use. During that time, several significant changes have been made in this material to furnish greater versatility or to improve the quality of results. When first introduced in 19501 as Type 5247, the film was balanced for daylight or for the high-intensity carbon arc illumination then currently used for studio color production. Changes in studio lighting practice dictated the need for a film balanced for tungsten illumination. Accordingly, Type 5248 was designed to meet this need and was introduced in 1953.2 Along with this change, improvements were also realized in the way of reduced granularity and better color reproduction.
A higher speed film, Type 5250, was introduced in 1959.3 Now widely used for color productions, this film provided twice as much speed as the earlier product without significant sacrifices in granularity or sharpness. This speed increase made possible greater economy in set lighting, or permitted exposures to be made under more difficult lighting conditions or at smaller apertures for greater depth of field. Changes were also incorporated in the design of the film to furnish further improvement in color reproduction.
Shortly after the introduction of Type 5250, work was started with the aim of making an even better product. In the light of existing technology, it was felt that the development program could proceed along either of two lines, namely, toward an increase in speed with no change in granularity or, alternatively, toward a reduction in granularity with no change in speed. Considering the high magnifications employed in present-day widescreen pictures, it was concluded that reduced granularity would offer a greater advantage than higher speed to most users of the product. Efforts were therefore concentrated in this direction, resulting in the film to be described.
Structure of Film
The new film is designated as Eastman Color Negative Film, Type 5251. Its structure is fundamentally the same as that of its predecessors and only a brief description will therefore be given. Adjacent to the safety support is the blue- and red-sensitive emulsion layer. Just above this layer is the blue- and green-sensitive emulsion layer. Next in order is the yellow filter layer, which serves to prevent blue light from reaching the bottom layers during exposure. Above the yellow filter layer is the blue-sensitive emulsion layer. A jet antihalation coating is contained on the reverse side of the support as for other Eastman Color Films.
The emulsion layers contain appropriate dye-forming couplers which yield the proper color images in each layer upon development. As with the earlier materials, the couplers contained in the two bottom emulsion layers are themselves colored. The unused portions of these couplers remain in the film after processing and serve as masks to correct for the unwanted absorptions of the process dyes. A description of this masking scheme has been given in an earlier paper.1
The Type 5251 film is color balanced, in the same manner as Type 5250, for studio tungsten lamps operating at a color temperature of 3200 K. The film may of course be used under daylight illumination conditions using the Kodak Wratten Filter No. 85 over the camera lens. Other illuminants may also be employed using the same conversion filters as recommended for Type 5250.
Since Type 5251 is designed to have the same speed as that of Type 5250, those who are familiar with exposure of the latter may continue to use their preferred exposure index values. For those who may be using Eastman Color Negative Film for the first time, it is recommended that an exposure series be made at settings above and below those called for when using an exposure index of 50 for tungsten and 32 for daylight with the No. 85 filter. Higher or lower values may be found desirable depending on the individual meters and method of use.
Incident Illumination Values
Most cameramen will prefer to measure the incident illumination. In this case, the basic level of 100 ft-c for an aperture of f/2.0 and 1/50 sec is generally satisfactory.
The new film exhibits slightly different spectral sensitivities for the three emulsion layers than those shown by Type 5250, as illustrated in Fig. 1. With reference to the yellow-forming layer, the peak sensitivity is at the same wavelength as Type 5250. The sensitivity of the magenta-forming layer extends slightly further toward the short wavelength region but is about the same as Type 5250 in the long wavelength region. The cyan-forming layer shows a slightly broader sensitivity, extending further toward both shorter and longer wavelength regions than the corresponding layer of Type 5250.
These differences may cause certain object colors to be rendered differently with Type 5251 than with Type 5250.
However, it is reasonable to expect that most studios will want to make wardrobe and makeup tests before proceeding with production work using the new film.
The processing solutions and procedures are the same as those recommended for Type 5250.4 One point which might be mentioned is that the clearing time in the first and second fixing baths is slightly longer for the new film than for the old film. The recommended total fixing time of 4 min is entirely adequate, however, for Type 5251.
As with Type 5250, either the dichromate or ferricyanide bleach solutions may be used. It has previously been reported that a dichromate bleach produces an edge effect as a result of its tanning action.5 This effect shows up in the print as a slight improvement in sharpness. A dichromate bleach also causes the surface layer of the negative film to be rougher in character. If the printer illumination conditions are specular in nature, this rougher surface may appear as increased graininess in the print.6 These same effects occur with Type 5251, but to a lesser degree.
A ferricyanide bleach does not produce the rough surface and will therefore be preferred for minimum graininess in the prints. Another advantage to be gained by the use of a ferricyanide bleach is that it may be regenerated easily with sodium or potassium persulfate.*
The recommendations for handling and storage of raw stock and processed negative are the same as those stated for Type 5250.4
In making exposures for process control, the same procedures as those recommended for the old product may be used.4 The sensitometric strips are measured on a photoelectric densitometer which measures the integral densities to red, green and blue light. It is preferable to use filters designed to give readings representing densities which the negative presents to the print film. Typical curves plotted from such readings are shown in Fig. 2.
It will be noted from Fig. 2 that the sensitometric curves for Type 5251 exhibit better linearity and somewhat greater latitude than those for Type 5250, especially for the blue- and green-sensitive emulsion layers. This results in better overall tone reproduction, particularly with reference to the neutral scale.
It is to be noted that the minimum blue density of Type 5251 has been adjusted so as to be very close to that of Type 5250. The minimum green density for Type 5251 is slightly higher than that found for the old film. The significance of these differences will be apparent in the description of the printing characteristics.
Spectral Density Curves
The couplers used for the new film yield the same dyes upon development as are produced in the Type 5250 film. However, a cyan coupler having a slightly different original color, and hence slightly different masking properties, has been used in the new film. Slight differences in the equivalent dye density curves for the new and old films are therefore noted, as illustrated in Fig. 3.** The only practical significance of this change is seen in the printing characteristics discussed below.
Granularity and Graininess
The reduced granularity obtained in the new negative film results in lowered graininess as observed in the projected print. This improvement will be of particular benefit where high magnifications are employed, such as in wide-screen and drive-in theater presentations. It is also of special interest to note that the improvement in granularity of the original negative material is carried through to the final print even when duplicating stages are employed. In fact, prints made from duplicate negatives derived from Type 5251 are very nearly comparable from a graininess standpoint to prints made directly from Type 5250 negatives. This result is most gratifying since any enhancement of the quality of portions of a production involving special effects, where duplicate negatives are employed, has constantly been sought. Distinct benefits will also be realized in instances where release prints are made from duplicate negatives of the complete production, as is often done in overseas countries.
One question which naturally arises is what if any differences exist in the printing characteristics of Type 5251 relative to the older film. To answer this question completely, it should be noted that the new negative film is tailored primarily for use with an improved Eastman Color Print Film, Types 5385 and 7385 discussed in a companion paper.7 For some months ahead, laboratories may therefore be called upon to print both old and new type negatives on either the old or new type print films, and any one of four different film combinations will be encountered. The printing characteristics for both subtractive and additive printing systems are therefore presented with these facts in mind.
The data shown in Table I illustrate the filter corrections required for combinations of the new and old type negative and print films for a typical subtractive-type printer. The data are derived from tests made on a modified Bell & Howell printer equipped with a 500-w lamp, operating at 100 v, a Pittsburgh Heat Absorbing Glass No. 2043 and a diffuser. For ease of comparison, all data shown are for operation of the printer at 90 ft/min and at diaphragm 11. Modifications of the filter packs will be required for other printing conditions and for different emulsion numbers of print stock.
The data shown in table II are derived from tests made on a modified Bell & Howell printer equipped for additive color printing. The printer used employs three 1000-w projection lamps operating at voltages ranging from 50 to 100 (to control intensity and color balance) and with suitable dichroic mirrors and filters. The printer speed is 90 ft/min.
For convenience of comparison, it is assumed that the color balance conditions for printing Type 5250 negatives onto Type 5382 print stock are known; therefore, the red, green and blue exposure level is arbitrarily given as 100% for each. For other negative and print film combinations, the corresponding log E or percentage changes in red, green and blue exposure levels are given.
As an example of the use of the data given in Table II, suppose that the printer calls for lamp voltages of 88 for the red, i77 or the green and 59 for the blue when printing Type 5250 onto Type 5382. If it is now desired to print Type 5251 onto Type 5382, reference to the table below shows that no increase is needed in the red and blue exposure but the green exposure must be increased by 63% or by a log value of 0.20. Reference to the calibration chart for the particular printer used shows that for a log exposure increase of 0.20 the voltage on the green lamp must be increased by 12. The new voltages are therefore 88, 89 and 59, respectively.
The requirements for making the most satisfactory color duplicate negatives from Type 5251 are the same as those stated in a previous paper when working from Type 5250 negatives.8 In general, the individual operations are carried out in such a way as to utilize the linear portions of the curves of the color duplicating material used for preparing the color master positive and duplicate negative. Very briefly, this consists of first measuring the densities of the significant highlight and shadow areas of the scene and correlating these with reference steps of a gray scale. These steps are plotted on the control curves for the original negative.
Secondly, the color master positive is exposed in such manner as to position the significant scene information on the linear portions of the blue, green and red curves of the duplicating film.
Thirdly, the color duplicate negative is exposed so that the important scene information is again positioned on the linear portions of the curves of the duplicating film.
If the operations are carried out in this manner, the duplicate negative will have a higher overall density level than the original negative and will require different exposure and color balance adjustments in printing. Our experience in making duplicate negatives which can intercut with original footage with no intensity and color timing indicates that an undesirable loss in contrast and color quality occurs. Irrespective of whether the duplicating procedures used depart from these ideal conditions, duplicate negatives made from Type 5251 should yield a final print quality at least equal to or, in most cases, better than that derived from Type 5250 negatives because of the better linearity and lower granularity characteristics of the new negative film.
Note: At the conclusion of the paper, a demonstration film was projected showing the quality of prints made directly from the Type 5250 and Type 5251 negatives, alternately with the same subjects, and also using split-frame projection. Prints were also shown which were prepared from duplicate negatives derived from the old and new type negatives. Finally, a 16mm reduction print prepared from the new-type negative was projected.
* U.S. Patent 2,944,895, July 12, 1960, assigned to Eastman Kodak Co.
** The equivalent dyes represent the contribution of the individual layers to the observed variations in spectral density obtained by varying exposures on normally manufactured and processed film. For a further explanation, see “Equivalent dyes for color films with incorporated couplers” by A. J. Sant, Phot. Sci. & Eng., 5, No. 3, May–June 1961.
1 W. T. Hanson, Jr., “Color negative and color positive film for motion-picture use,” Jour. SMPTE, 58: 223–238, Mar. 1952.
2 W. T. Hanson, Jr., and W. I. Kisner, “Improved color films for color motion-picture production,” Jour. SMPTE, 61: 667–701, Dec. 1953.
3 M. L. Dundon and D. M. Zwick, “A high-speed color negative film,” Jour. SMPTE, 68: 735–738, Nov. 1959.
4 “Production of motion pictures in color using Eastman color films,” Motion Picture Film Dept., Eastman Kodak Co., Rochester 4, N.Y. Available upon request.
5 H. J. Bello and D. M. Zwick, “An edge effect in color films caused by distortion of the developed image when a hardening bleach is used,” Phot. Sci. & Eng., 3: No. 5, 221–224, Sept.–Oct. 1959.
6 D. M. Zwick, “How color negative film surface characteristics affect picture quality,” Jour. SMPTE, 71: 15–20, Jan. 1962.
7 W. I. Kisner, “A higher speed color print film,” Jour. SMPTE, 71: 779–781, Oct. 1962.
8 R. O. Gale and W. I. Kisner, “Techniques in color duplication,” Jour. SMPTE, 68: 874–881, Dec. 1960.”
(Kisner, W. I. (1962): A New Color Negative Film for Better Picture Quality. In: Journal of the Society of Motion Picture and Television Engineers, 71,10, pp. 776–779.)
“Goskino began to import a limited amount of Eastman Kodak negative. Bought with western currency through European intermediaries, it was strictly rationed and highly prized (Tarkovskaya 1990). The choice of which projects were allocated imported stock was inevitably political. Kodak was a mark of favour, reserved for Goskino’s preferred projects. The films that used it were, most often, those granted ‘highest category’ and ‘first category’ status, flagship films including Tretya molodos/Nights of Farewell (Jean Dreville and Isaak Menaker, 1966) and Voyna i mir/War and Peace (Sergei Bondarchuk, 1968).2 Less privileged directors had two options: black-and-white or Sovcolor.
2 There were five categories: highest, first, second, third, and fourth (Golovskoy 1986: 47).
Golovskoy, V. (1986), Behind the Soviet Screen: The Motion-Picture Industry in the USSR 1972-1982, Ann Arbor: Ardis.”
(Misek, Richard (2007): “Last of the Kodak”. Andrei Tarkovsky’s Struggle with Colour. In: Everett, Wendy (ed.): Questions of Colour in Cinema. From Paintbrush to Pixel. Oxford: Peter Lang, pp. 161–176, on p. 163.)
“Tarkovsky’s response to the limited availability of colour was to attempt to obtain as much Eastman Kodak as he could.”
(Misek, Richard (2007): “Last of the Kodak”. Andrei Tarkovsky’s Struggle with Colour. In: Everett, Wendy (ed.): Questions of Colour in Cinema. From Paintbrush to Pixel. Oxford: Peter Lang, pp. 161–176, on p. 164.)
“Unfortunately, even when a film was granted imported stock, the amount allocated was often inadequate. Part-way through filming Solaris, the Kodak ran out (Tarkovsky 1994: 39).3 Such shortfalls were routine at the time, and the most common solution was to mix Kodak and Sovcolor. Tarkovsky’s response was more unusual. He initially decided to film the remaining scenes in black-and-white (Tarkovsky 1994: 42).4 In the event, some of the scenes towards the end of the project’s shooting schedule were indeed filmed in black-and-white, but others were filmed in colour. Was Tarkovsky able to negotiate more Kodak? It is impossible to be sure. However, what the available evidence does make apparent is the fact that the inclusion of black-and-white sequences in Solaris was at least partially dependent on supply-side economics.
Supply-side economics played an equally important role in Tarkovsky’s next film. In Mirror, he was again faced with the problem of inadequate colour stock, as is highlighted in the following diary entry:
On Monday 9th, I went to see Pavlyohok. He’s an unpleasant, coarse, louche character. He was bawling at Erica M. and Karayev and trying to drive a wedge between them. He and Yermash (or rather, Yermash) had been given a direction. The result, we heard yesterday, is that we are being given both money – 622,000 roubles – and 7,500 metres of Kodak film. That means up to three takes. We shall have to get the other 3,000 metres x 4 from Konoplyov (Tarkovsky 1994: 77).
Tarkovsky’s terse writing style, combined with his casual insider’s references to the Soviet film industry, often make his diaries difficult to decipher. Who are all these people? What role does each of them play? What is the significance of the parenthesis (it seems to suggest a discrepancy between the official and actual responsibilities of Boris Pavlyonok, Deputy Chairman of Goskino, 1973-1985)? Despite these uncertainties, the above entry provides a crucial piece of information. It suggests that the ratio of Kodak to ‘other’ film stock available for the production was about 2.5:1. Given a final running time of 104 minutes, and assuming an editing strategy in which the initial ratio of Kodak to ‘other’ stock was maintained, the final film should comprise about 74 minutes of Kodak and 30 minutes of ‘other’ stock. In fact, 71 minutes of Mirror are in colour and 33 minutes are in black-and-white. This startlingly close correlation strongly suggests that the ‘other’ stock that Tarkovsky used was black-and-white.5 In other words, as in Solaris, the extent to which Tarkovsky mixed black-and-white and colour was closely related to the simple logistical question of what film stock was available.
In Stalker, Tarkovsky again secured the use of Kodak, but this time he was given a poisoned chalice. Something went badly wrong when the stock was processed. Accounts of what precisely happened at the Mosfilm laboratories and why, and of precisely how much footage was destroyed, differ. Regardless of the details, the problem was serious enough to halt production. Amazingly, Tarkovsky was able to resurrect the project and start filming the following year with a new script, a new cinematographer, and a new art director (Johnson and Petrie 1994: 138). By this time, however, ‘the last of the Kodak’ was exhausted and no more was forthcoming (Tarkovsky 1994: 146). One might imagine that at this point Tarkovsky would have reshot the film entirely in black-and-white, as Goskino wanted him to do, or entirely in Sovcolor. In fact, even though he was no longer faced with the dilemma of how to make optimal use of a limited supply of Kodak, he again chose to mix colour and black-and-white. In the absence of an obvious economic motivation for this mixture, one must conclude that the choice was made for reasons internal to the film. In other words, in Stalker, Tarkovsky used black-and-white not because he had to but because he wanted to.
3 The shortfall in colour stock is confirmed by Mikhail Romadin, Solarises Art Director, in a recent DVD interview (Solaris Special Edition DVD, The Criterion Collection, 2002).
4 Though Tarkovsky’s mixture of colour and black-and-white was unusual, it was not unique. Examples of contemporaneous Soviet films that mixed the two formats – presumably also for reasons influenced by supply-side economics – include Dyadya Vanya (Andrei Konchalovsky, 1970), The Dawns Here are Quiet (Stanislav Rostotsky, 1972), and Only Old Men are Going to Battle (Leonid Bykov, 1973).
5 The small discrepancy between these statistics can be explained by the fact that no production goes entirely according to plan. It is inconceivable that there were not occasional divergences from Tarkovsky’s anticipated 3:1/4:1 shooting ratios: It is also inconceivable that decisions made during the editing process did not also influence the film’s final ratio of colour to black-and-white.
Johnson, V. and G. Petrie (1994), The Films of Andrei Tarkovsky: A Visual Fugue, Bloomington: Indiana University Press.
Tarkovsky, A. (1994), Time Within Time, The Diaries 1970-1986, London: Faber and Faber.”
(Misek, Richard (2007): “Last of the Kodak”. Andrei Tarkovsky’s Struggle with Colour. In: Everett, Wendy (ed.): Questions of Colour in Cinema. From Paintbrush to Pixel. Oxford: Peter Lang, pp. 161–176, on pp. 165–167.)
“L’Eastmancolor a été la première pellicule couleur monopack à conquérir le marché: elle apparaît en 1950 avec une sensibilité de 16 ASA, passe à 25 ASA en 1952, à 50 ASA en 1959, et à 100 ASA seulement en 1968. Il faudra attendre 1981 pour que Fuji mette sur le marché européen la révolutionnaire Fuji 250 qui titre 250 ASA en usage normal et permettra de tourner enfin en 35 mm couleur dans les conditions qui étaient celles du noir-et-blanc au cours des années soixante9. Kodak suivra de très près en commercialisant dès 1982 un négatif Eastman à 250 ASA.
9 Grâce à cette pellicule, Neige, de Juliet Berto et Jean-Henri Roger, a été tourné en grande partie la nuit, à Pigalle, pratiquement sans éclairage, comme avait pu le faire Godard en noir-et-blanc au cours des années soixante.”
(Bergala, Alain (1995): La couleur, la Nouvelle Vague et ses maîtres des années cinquante. In: Jacques Aumont (ed.): La Couleur en cinema. Milan: Mazzotta, pp. 126–136, on p. 129.) (in French)
“Il peut sembler étrange, aujourd’hui, que l’on rassemble ainsi, dans une commune esthétique de la couleur, au cours des années cinquante, des films et des cinéastes européens et américains. C’est sans doute lié au fait d’une certaine homogénéité du rendu des couleurs au cours de ces années cinquante. Même si c’est la décennie où le déclinant Technicolor cède en douceur la place à l’Eastmancolor – lequel n’a pas encore la transparence qui sera son apanage au cours des années soixante-dix, mais un rendu des couleurs assez proche du Technicolor au point qu’il est parfois difficile de distinguer si certains de ces films sont tournés avec l’une ou l’autre de ces pellicules17 – les lourdes contraintes d’éclairages (nécessitant des arcs puissants) sont presque les mêmes, en intérieur, pour tous les films en couleur, à Hollywood ou en Europe, ce qui a tendance à homogénéiser le traitement de la couleur de ces films, quelles que soient les différences d’esprit et de production qui les distinguent par ailleurs – même si l’Eastmancolor est déjà deux fois plus sensible que le Technicolor (50 contre 20 ASA). Autre facteur d’homogénéisation, il semble qu’au cours des années cinquante, les chefs opérateurs n’aient jamais pris le risque d’une prise de vues avec un diaphragme inférieur à 4,5, pour ne pas mettre le point en péril. Jean Rouch prétend que, en 16 mm et à 25 ASA, il tournait dès la fin des années cinquante à pleine ouverture (avec un objectif ouvrant à 0,95), ce qui équivalait à utiliser une pellicule de 200 ASA à l’ouverture maximale de l’époque en 35 mm.
17 Renoir, par exemple, tourne Le Fleuve (1951), Le Carrosse d’or (1952), French Cancan (1955), Elena et les hommes (1956) en Technicolor, et Le Déjeuner sur l’herbe (1959) en Eastmancolor.”
(Bergala, Alain (1995): La couleur, la Nouvelle Vague et ses maîtres des années cinquante. In: Jacques Aumont (ed.): La Couleur en cinema. Milan: Mazzotta, pp. 126–136, on p. 131.) (in French)
“In May ’47 the US Government filed an anti-trust suit against Technicolor and Eastman Kodak, charging restraint of trade and monopolization of professional color cinematography. The court proceedings, as to Eastman, ended in ’48 with a consent decree, and, as to Technicolor, in the same way in ’50.
The Government’s decrees, and the background of the proceedings, can be found in two papers and an appendix written by George S. Frost and S. Chesterfield Oppenheim for the 1960 spring and summer issues of The Patent, Trademark and Copyright Journal of Research and Education, which is published by George Washington University. I quote from their summary: “It is concluded in the present report that none of the compulsory licensing decree provisions has had substantial effect on the professional color motion picture industry, that the key development since the decrees (Eastmancolor) occurred without regard to the decrees, and that research and development in the industry has a much greater effect on market competition than the anti-trust decree provisions.”
In ’49 Eastman Color single-strip negative and printing film stock were introduced for professional motion pictures. It was the most far-reaching breakthrough since Technicolor’s three-strip process in ’32. Within five years the Technicolor camera was a thing of the past. Eastman Color negative, in a standard black-&-white camera, was the universal tool. The WarnerColor, MetroColor, De Luxe Color, Pathe Color, etc., that appeared in film ads and on credit titles are all Eastman Color stock. The different “names” refer only to the studio and/or processing plant that did the developing and printing.
The perfection of Eastman Color, and the advent of CinemaScope (’53), radically altered Technicolor’s entire modus operandi. It began to do color negative processing and color positive printing, and to “diversify”. It also promoted such wide-screen developments as Technirama and Technivision-70.”
(Behlmer, Rudy (1964): Technicolor. In: Films in Review, 15,6, pp. 333–351, on pp. 350–351.)
“When first introduced in 1950,1 Eastman Color Print Film, Type 5381 was intended to serve as a companion to Eastman Color Negative Film, Type 5247. Although it was recognized as a distinct accomplishment for its time, it was apparent that a product having better definition would be very desirable. Work was therefore continued with this aim in view.
In 1953, an improved product was described2 which provided not only better definition but also better color reproduction because of certain changes made in the color couplers. As a matter of fact, the definition was sufficiently improved to permit furnishing the product in both 35mm and 16mm widths designated as Types 5382 and 7382, respectively.
During the period from 1953 to 1959, a gradual increase in the use of the 16mm product, Type 7382, was noted both for direct reduction printing from 35mm color negatives as well as for contact printing from 16mm color internegatives. For such work, it was evident that a print film of higher definition would be of distinct benefit. By 1959, it appeared feasible to in corporate modifications in the Type 7382 to give some increase in definition and it was therefore replaced by Type 7383. Research and development work continued with the aim of effecting still further improvements.
The road ahead looked favorable indeed, with all signs pointing to a bright horizon. It seemed feasible to make a product which would have the same granularity as the current materials, definition characteristics similar to those of Type 7383 and higher effective speed as well. In addition, it appeared possible to make use of two new couplers which would result in better color reproduction.
Structure of Film
The new film is designated as Eastman Color Print Film, Types 5385 and 7385. The 16mm product is cut from the same parent rolls as might also be used for the 35mm product. The structure is basically the same as that of the older material with the blue-sensitive emulsion located nearest to the support, the red-sensitive emulsion just above it, and the green-sensitive emulsion in the uppermost position. Colorless couplers which produce the appropriate dye images upon development are contained in each emulsion layer. A jet antihalation backing is located on the reverse side of the support.
New magenta and yellow couplers have been utilized for the new film to give improved color reproduction, as discussed later.
The spectral sensitivity characteristics of the new film are very similar to those of the older product. However, it will be noted by reference to the separate ordinate scales for the solid and dotted curves of Fig. 1, that the overall sensitivity of Type 5385 is considerably higher than that of Type 5382. From a practical standpoint, this amounts to an effective printing speed increase of about four times. This speed increase should be of great value to laboratories in the way of providing faster printing speeds and greater production output or in giving greater flexibility in instances where printers may have limited intensity levels and printing range.
It is also of interest to note that the minimum sensitivity of Type 5385 occurs at a wavelength of about 588 mµ which is near the wavelength of the sodium lines (589 mµ). The minimum sensitivity of Type 5382 also occurs near this wavelength. These facts are important in relation to the selection of proper safelights for handling the new film, as mentioned later.
Spectral Density Curves
As shown in Fig. 2, the spectral density curves for the new film are slightly different from those for Type 5382 because of the new yellow and magenta couplers used for the former. The peak absorbance of the magenta dye is located further toward the short wavelength region by about 15 mµ. The peak absorbance for the yellow dye is also located slightly further toward the short wavelength region by about 5 mµ.
The practical significance of the modifications in the dye system can be seen by viewing actual samples of the processed films which have been exposed with suitable filters to give red, green, blue, yellow and neutral images. The Type 5385 film gives brighter and less magenta reds, more saturated greens, less cyan blues, and yellows which are less orange in hue than those given by the older film. A still better appreciation of these differences will be obtained in viewing actual pictures of various subjects. Better reproduction of blue skies and foliage, of natural fruits and vegetables, of flesh tones and certain paints and fabrics will be observed.
In exposing sensitometric strips for control purposes, it is usually most convenient to adjust the filter pack in the sensitometer in such a manner as to obtain a neutral scale. Since the speed relationships of the three emulsion layers for Type 5385 are slightly different from those of Type 5382, a different filter pack will be required. Also, since the overall speed of the Type 5385 is about four times that of Type 5382, some sensitometers may call for a reduction in the illumination level, such as by use of a suitable neutral density filter.
The integral densities of the control strips are measured in the same way as for Type 5382 strips. The D-log E curves for the new and old print films, shown in Fig. 3, are very similar with only slight differences seen in the toe portions. It should be noted that the curves shown in the illustration have been positioned in such a way as to illustrate only the relative curve shapes and do not, therefore, show the speed relationships for the separate emulsion layers or for the two products. These relationships will be apparent upon examination of the printer balance requirements for the new film as discussed in a subsequent paragraph.
The fourfold increase in speed noted for the new film naturally calls for a reconsideration of the safelight conditions. For the older film, a Kodak Safelight Filter, Wratten Series 8 has been recommended. This same filter could also be used for the Type 5385 film if the intensity is reduced by at least four times. With the reduced level, the visibility is likewise reduced and some laboratories may therefore prefer to use a Wratten Series 3 instead of a Series 8 because of the higher sensitivity of the eye to green light. Consideration might also be given to the use of sodium vapor lamps with additional filters (Kodak Wratten Filters No. 23A and No. 57) to remove radiation of unwanted wavelengths and a neutral density filter to reduce the intensity to a safe level. In any case, each laboratory should make careful safelight tests before proceeding with production work. It is also worthwhile to consider installation of safelights along the floor paths or in locations where the safelight illumination cannot reach the film directly. It is surprising to find how much higher the safe illumination level can be in this kind of an arrangement.
The solutions, processing steps and treatment times are the same as those recommended for Type 5382.
Of particular interest to laboratories is the subject of printer speed and color balance conditions to be used with the new film. The higher effective speed of the new film will permit a reduction in overall printer light intensity or, alternatively, faster printer operation and greater production output. As mentioned earlier, the printers in use in some laboratories have had a marginal light-intensity level, and this condition has limited the printing range when using the old type of print film. The higher speed print film will, therefore, provide greater flexibility of operation in such cases.
It is expected that by the time commercial quantities of Types 5385 and 7385 are available a substantial percentage of negative footage received for printing will be on the new Type 5251 negative film, described in a companion paper.3 However, laboratories will be called upon to make prints on the new stock from Type 5250 negatives of some of the recent and also some of the older productions. It is likely too that some laboratories will have remaining stocks of the older type of print film which they will want to use. Thus, four different negative and print film combinations must be handled for some time in the future. The printer color balance conditions are somewhat different for each situation and suggested balance conditions for both subtractive and additive printers have been given in the companion paper referred to above.
In printing the soundtrack, the exposure should be confined to the top two emulsion layers as is now recommended for printing onto Type 5382. A suitable filter combination for this purpose consists of Kodak Wratten Filters No. 2B and No. 12. The optimum sound negative and print densities should be determined by making listening tests or, if equipment is available, by making cross-modulation or intermodulation tests. The following data determined under one set of laboratory conditions may be helpful as a guide.
Printing Color Internegatives Onto Color Print Film
The new print film will of course find immediate application in preparing prints from color internegatives derived from reversal color originals such as those made on Ektachrome Commercial Film, Type 7255. In this production system, the internegative is made on Color Internegative Film, Type 5270 or 7270.4 As in other duplication procedures, it is recommended that the essential scene information be transferred to the internegative material in such a way as to utilize the linear portions of the D-log E curves for the best color reproduction.
In printing the color internegative to the new print film, certain color balance and intensity adjustments will also be required relative to those used with the older type print material (Type 7383). The following data determined from tests made on a typical subtractive-type printer will serve as a guide in arriving at the correct balance conditions.
Storage of Raw Stock and Processed Film
The recommended conditions for storage of raw stock are the same as those previously given for the older print materials. For periods of time up to three months raw stock may be stored at temperatures not exceeding 50 F without significant changes in its sensitometric properties. If stocks are to be kept for longer periods of time it is advisable to store them at temperatures lower than 50 F to minimize any changes in sensitometric properties. Upon removal from storage, ample time should be allowed for the film to reach equilibrium with the workroom conditions before the tape is removed from the can in order to avoid condensation of moisture on the cold film from the atmosphere.
Note: At the conclusion of the paper, demonstration prints were projected to illustrate the results obtained with the new print film and also to show the overall improvement provided by the combined use of the new color negative and print films.
1 W. T. Hanson, Jr., “Color negative and color positive film for motion picture use,” Jour. SMPTE, 58: 223–238, Mar. 1952.
2 W. T. Hanson, Jr., and W. I. Kisner, “Improved color films for color motion picture production,” Jour. SMPTE, 61: 667–701, Dec. 1953.
3 W. I. Kisner, “A new color negative film for better picture quality,” Jour. SMPTE, 71: 776–779, Oct. 1962.
4 D. M. Zwick, H. J. Bello and C. E. Osborne, “A 16mm color internegative film for use in color motion-picture photography,” Jour. SMPTE, 65: 426–428, Aug. 1956.”
(Kisner, W. I. (1962): A Higher Speed Color Print Film. In: Journal of the Society of Motion Picture and Television Engineers, 71,10, pp. 779–781.)
Racconti romani (ITA 1955, Gianni Franciolini)
“Uno dei film che aprì la strada alla commedia bozzettistica a colori fu Racconti romani, girato in Eastmancolor. Il film propone un uso piuttosto distillato del colore, all’interno delle inquadrature espanse del Cinemascope. Le riprese in esterni reali vi assumono complessivamente un ruolo più importante rispetto ai generi fin qui presi in considerazione. Nella sequenza dei titoli di testa, ad esempio, le scritte scorrono su una serie di vedute di edifici e monumenti capitolini, che si configurano sul piano cromatico per la prevalenza di una gamma tra il verde scuro degli alberi, il colore plumbeo del cielo e le sfumature di grigio e marrone degli elementi urbanistici. La presenza di tinte neutre o smorzate veniva spesso considerata più consona alle rappresentazioni ascrivibili alle categorie del realismo e il film si conformava a questa stessa idea.
Questa sorta di compendio iconografico della città di Roma innestava la novità del colore su immagini molto note. Lo spettatore non doveva percepire uno scarto tra i luoghi rappresentati sullo schermo e i colori che lui stesso era portato ad attribuire a quegli stessi luoghi. Anzi, era proprio nella possibilità di vederli nei loro presunti colori naturali che il patto con lo spettatore poteva rinsaldarsi. Quest’ultimo veniva invitato alla visione cromatica di monumenti che altri film e immagini gli avevano presentato nella forma prevalente del bianco e nero. A garantire la continuità della clausola realistica contribuiva il gioco delle variazioni tonali, che, come nel bianco e nero, continuava a definire profondità e relazioni spaziali all’interno delle inquadrature.
In alcune sequenze, inoltre, il colore acquista il ruolo di moltiplicatore visivo delle relazioni comportamentali e dei moti desideranti che si attivano tra i personaggi. La giovane Ines, cameriera nel bar abitualmente frequentato dal gruppo di amici al centro del film, agisce in un ambiente caratterizzato da una più marcata presenza cromatica: a colorare l’interno sono infatti le vernici del bancone e degli scaffali, il registratore di cassa, le bottigliette multicolore, il calcio balilla. Alla prima apparizione nel bar, la ragazza indossa un abito bicolore, azzurro chiaro nella parte superiore e bianco a pois multicolori in quella inferiore; i quattro amici, per contrasto, non escono dalla gamma dei bianchi e dei blu-grigi. Circolando nelle inquadrature girate in questo ambiente, il colore contribuisce in maniera decisiva a erotizzare lo spazio, a partecipare visivamente dell’euforia dei personaggi maschili. Anche altre sequenze sono segnate dalla ricorrenza di colori portati da oggetti, manifesti pubblicitari e insegne commerciali: la disseminazione di questi elementi tra le immagini corrisponde alla domanda di eros, svago, consumo e spensieratezza che i giovani protagonisti sono ansiosi di formulare alle proprie vite.
Assieme ad altri film coevi, Racconti romani favorì il processo di normalizzazione del colore sugli schermi. La sua presenza veniva infine sottratta a investimenti puramente artificiali e spettacolari per essere inscritta in un orizzonte realistico. Per giustificare la scelta del colore, non era più necessario cercare soccorso nelle tavole di un palcoscenico, nella tela di un dipinto o in un libro di storia; era sufficiente volgere lo sguardo verso il mondo circostante, per cogliere gli indizi cromatici di cui appariva disseminato.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 210–212.) (in Italian)
Mon oncle (FRA/ITA 1958, Jacques Tati)
“Tati, ad esempio, nei suoi film a colori ripensò cromaticamente l’universo diegetico dei suoi personaggi, trasformandolo in un formidabile campionario di oggetti colorati20. Mio zio (Mon oncle, 1958) fa dell’automobile uno dei tanti mezzi tecnologici presenti nel quotidiano del ricco industriale Arpel, ansioso di sperimentare tutto quanto si configuri come nuovo e funzionale, dall’automazione dei dispositivi alla conversione ai nuovi oggetti di design e ai più aggiornati criteri costruttivi. In una delle sequenze iniziali, il personaggio esce di casa con una elegante berlina grigia e bianca, mescolandosi visivamente con le altre automobili sulla strada, accomunate dagli stessi colori. Il montaggio compone una successione di immagini impostate sulla dominanza di tinte metalliche, metallizzate o cementificate, cui si conformano, oltre alle auto, i muri della casa, gli abiti, l’asfalto, l’esterno della scuola e dell’azienda. Già attraverso una simile ricorrenza del grigio il film riesce a marcare una distanza incolmabile rispetto alla visione in bianco e nero. Tutti questi grigi piatti e a campiture nette, attraversati da vari riflessi cromatici, con la loro brillantezza od opacità, levigatezza o granulosità, hanno ormai poco a che fare con i grigi delle immagini in bianco e nero. La relazione significante tra grigio e colore viene spostata su un piano più raffinato, meno scontato e banale rispetto a Racconti romani.
Il film apre al grigio una nuova visibilità, sottraendolo al dominio dell’incolore per riqualificarlo a dimensione cromatica costitutiva del mondo tecnologico. Il grigio non è solo diventato un colore tra i colori, ma quello che, più di ogni altro, sembra capace di riconfigurare la fisionomia degli spazi e degli ambienti moderni, industriali e urbani. Il film esprime la particolare consapevolezza che soltanto la pellicola a colori possa restituire visivamente le qualità specifiche di questa nuova serie di grigi cromatici portata da nuovi materiali, plastiche, smalti, lacche, vernici, tinture, cementi, bitumi e prodotti sintetici21. In bianco e nero, tutte queste varianti pigmentarie del grigio avrebbero perso la loro forza. Tanto i grigi quanto i moderni colori consumistici sembrano così accomunati dai medesimi valori di superficie e di contorno, dalla disposizione ordinata, quasi chirurgica, dei rispettivi materiali e vernici. In tutte le sequenze dedicate al côté ultramoderno il film fa coabitare sistematicamente questi grigi con i colori più accesi. La serie degli oggetti domestici dispiega un ricco campionario di tinte industriali, frutto di un’ironica rilettura delle più recenti proposte del design: i divani verdi e le sedie con imbottiture rosa, gialle, verdi, l’ombrellino blu, gli abiti sgargianti della vicina di casa, la sciarpa a quadri rossi e neri di Arpel e del cane, il tosaerba, l’idrante e la palla rossi, la sedia a dondolo gialla. Sul finire del film, la gamma dei moderni colori è completata da una sgargiante berlina verde, lilla e rosa, che aggiorna ai dettami della moda del tempo il precedente modello grigio e bianco.
Tati rivelava dunque come la costruzione del nuovo visibile potesse sostanziarsi tanto nelle più moderne cromie, spinte al massimo grado di saturazione, quanto nelle nuove zone di grigio che, paradossalmente, proprio l’immagine a colori sembrava poter aprire allo sguardo.
20 Tati si era interessato al colore fin da Giorno di festa (Jour de fête, 1949) che aveva girato ricorrendo al sistema additivo lenticolare francese Thomsoncolor. Era però dovuto ricorrere alle riprese di sicurezza in bianco e nero per far approdare in sala il film: dai negativi del Thomsoncolor non si era stati capaci di tirare fuori alcunché di proiettabile e sarebbero occorsi oltre quarant’anni prima che l’operazione risultasse possibile. Giorno di festa sarebbe rimasto per decenni un film in bianco e nero (per approfondimenti, cfr. Ede 1995).
21 Il medesimo discorso può essere esteso al bianco: si pensi alle sequenze ambientate nella cucina.
Ede, Francois (1995), “Jour de fête” ou la couleur retrouvée, Cahiers du cinéma, Paris.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 223–224.) (in Italian)
“Nel 1955, la necessità di operare lavorazioni compatibili con la nuova pellicola a supporto unico Eastmancolor (1951) spinse la società ad aggiornare il vecchio sistema, trasformandolo in un metodo di stampa di copie ad alta qualità, applicabile a qualsiasi negativo (Technicolor 5).”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on p. 161.) (in Italian)
“L’introduzione del film a colori nelle cinematografie europee è infatti accelerata dalla progressiva disponibilità delle nuove tecnologie a supporto unico derivate dall’Agfacolor. Dopo la fine del conflitto, i brevetti del sistema tedesco diventano parte del bottino di guerra degli Alleati, e, nel giro di alcuni anni, tutte le principali industrie di materiali sensibili sono in grado di lanciare i propri marchi: Sovcolor e Anscocolor (1945), Gevacolor (1947), Ferraniacolor (1949), Eastmancolor (1951), Fujicolor (1955)3. Queste nuove pellicole impiegano un po’ di tempo prima di essere naturalizzate all’interno di contesti produttivi istituzionali, ma dopo alcuni anni di sperimentazioni, nella prima metà degli anni cinquanta esse sono in grado di dar corso a una produzione sempre meno occasionale di film a colori, come vedremo nella seconda parte del capitolo, esaminando le spinte progressive alla diffusione del colore nel cinema italiano.
3 Tutti questi sistemi sfruttano il principio della sintesi sottrattiva e la tecnologia del supporto unico a triplo strato di emulsione, introdotta con il Kodachrome e con l’Agfacolor risp. nel 1935 e nel 1936. Il negativo è composto da tre strati di emulsione, sensibili risp. al blu, al verde e al rosso. Ciascuno strato contiene elementi in grado di produrre, per sviluppo cromogeno, un colorante complementare a quello per cui lo strato stesso è sensibilizzato (risp. giallo, magenta e ciano). Il negativo così ottenuto presenta i colori complementari a quelli del soggetto. Una volta stampato su un positivo caratterizzato dalla medesima struttura, si ottiene un’immagine che riproduce i colori del soggetto. Nell’Eastmancolor (1951), questa stessa procedura viene resa più stabile grazie all’introduzione di un sistema di masking (che ottimizza le aree di assorbimento spettrale dei tre strati) e una diversa disposizione degli strati stessi (cfr. Limbacher 1969, pp. 53–72; Ryan 1977, pp. 113–212; Koshofer 1988, pp. 119–130).
Koshofer, Gert (1988), Color. Die Farben des Films, Wissenschaftsverlag Volker Spiess, Berlin.
Limbacher, James L. (1969), Four Aspects of the Film, Brussel & Brussel, New York.
Ryan, Roderick T. (1977), A History of Motion Picture Color Technology, Focal Press, London-New York.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 191–192.) (in Italian)
“Nel frattempo, la domanda di pellicola a colori veniva soddisfatta in misura sempre crescente dal negativo Eastmancolor, divenuto disponibile per le lavorazioni nazionali nel 1954. Nel giro di pochi anni, il Ferraniacolor perse il suo primato a tutto vantaggio della pellicola statunitense, che, grazie a supporti sempre più sensibili e stabili, aumentò il divario tecnologico con la concorrenza, garantendo anche costi sempre più competitivi18.
18 Il negativo Eastmancolor tipo 5247, introdotto nel 1951, dichiarava una sensibilità di 16 ASA. Il successivo tipo 5248 la elevava nel 1953 a 25 ASA. Nel 1959, il tipo 5250 raggiungeva i 50 ASA (cfr. “American Cinematographer” 1950; Foster 1953; Id. 1959). Si tratta di sensibilità ancora lontane da quelle garantite dal bianco e nero: già nel 1954, il negativo Kodak tri-X garantiva una sensibilità di 250 ASA. Introdotto in Italia in ritardo rispetto alle pellicole concorrenti, l’Eastmancolor fu impiegato per una quota variabile tra un terzo e un quinto della produzione italiana a colori tra il 1956 e il 1958, dopodiché divenne la pellicola a colori più utilizzata.
“American Cinematographer” (1950), New Eastman Color Film Tested by Hollywood Studios and Film Labs, in “American Cinematographer“, XXXI, n. 3, marzo 1950, pp. 95 e 102.
Foster, Frederick (1953), Eastman Negative-Positive Color Films for Motion Pictures, in “American Cinematographer“, XXXIV, n. 7, luglio 1953, p. 322.
Foster, Frederick (1959), A Faster Color Negative, in “American Cinematographer“, XL, n. 6, giugno 1959, pp. 364–365 e 368–370.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on p. 201.) (in Italian)
“In una direzione stilistica simile a quella evidenziata in Aida si mosse tra il 1953 e il 1954 la Lux, che assieme a Ponti e De Laurentiis elaborò una sorta di variante nazionale del prestige picture hollywoodiano in Technicolor. Alcune costose produzioni come Ulisse (1954), Attila (Francisci, 1954) e Teodora (Freda, 1953) rinvigorirono la tradizione nazionale del film storico-mitologico alla luce delle coeve strategie visionarie dei kolossal statunitensi, talvolta realizzati proprio a Cinecittà. I primi due furono girati con il vecchio Technicolor a tre negativi, il cui costo sarebbe risultato insostenibile per produzioni di piccolo o medio cabotaggio; il terzo con l’Eastmancolor, che richiedeva ancora di essere lavorato negli Stati Uniti. Dei kolossal di Hollywood questi film cercavano di imitare gli stili cromatici: l’assestarsi delle trame narrative su strutture relativamente più articolate imponeva anche in questo caso la ricerca di un bilanciamento tra spettacolo e colore.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 207–208.) (in Italian)
“Dalla contaminazione tra i modelli del genere storico mitologico e la formula produttiva del basso costo si originò sul finire del decennio il nuovo filone autoctono del peplum, che per qualche anno avrebbe consentito di massimizzare i profitti minimizzando i costi. L’avanzata del peplum corrispondeva all’inesorabile arretramento del Ferraniacolor sul fronte dei film a modesto impianto produttivo e alla parallela avanzata dell’Eastmancolor. A partire dal prototipo Le fatiche di Ercole (Francisci, 1958) l’appeal spettacolare del colore si abbinava con la visione espansa dei sistemi a schermo largo di recente introduzione31.
31 Cfr. Vitella 2007, p. 104.
Mazzei, Luca; Vitella, Federico (2007), Geometrie dello sguardo. Contributi allo studio dei formati nel cinema italiano, Carocci, Roma.
Vitella, Federico (2007), Una questione di standard. Il passaggio dall’Academy ai formati panoramici, in Mazzei/Vitella 2007, pp. 65–130.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on p. 209.) (in Italian)
“L’introduzione del negativo Eastmancolor nel 1951 – come si è visto – aveva spezzato il quasi ventennale monopolio della Technicolor, rendendo possibile girare a colori con una normale macchina da presa, senza dover noleggiare quella speciale brevettata dagli ingegneri di Kalmus. La quota dei film a colori hollywoodiani crebbe rapidamente dal quindici al cinquanta per cento tra il 1950 e il 1955.
L’industria di Hollywood fu tentata dal film tridimensionale e da quello a schermo largo, entrambi pensati in necessaria congiunzione con la presenza del colore5. Nel 1953, anno di introduzione del Cinemascope, il presidente della Fox dichiarò di voler produrre solo ed esclusivamente film a colori e con il nuovo sistema a schermo largo.
Intanto, i sistemi e le pellicole in uso si avviavano a garantire una più elevata uniformità degli standard e a colmare progressivamente il gap tecnologico ed economico che in precedenza aveva scavato un solco profondo tra il colore e il bianco e nero. Nel 1959, la Eastman lanciò un negativo a colori che raddoppiava la sensibilità della precedente versione, portandola a 50 ASA; nel 1968, furono raggiunti i 100 ASA9. Benché i supporti in bianco e nero avessero superato da tempo questi stessi livelli, anche i valori tecnici delle nuove pellicole a colori garantirono ai direttori della fotografia più ampi margini di scelta nelle riprese, rendendo possibile girare in esterni con luce naturale e in interni con mezzi illuminanti più leggeri10.
5 Tra il 1952 e il 1953, il 3D si trasformò rapidamente da passatempo per amatori a oggetto d’interesse delle major. Il suo impatto fu tuttavia di corto respiro e già nel 1954 la moda del 3D era terminata. I produttori gli preferirono la soluzione dello schermo largo, sperimentando l’opzione di lusso del Cinerama o quella più semplice del Cinemascope o della lunga serie di imitazioni e varianti che ne derivarono (cfr. Lev 2003, pp. 107–125). In concomitanza con la diffusione dello schermo largo, si registrò anche una variazione del formato Academy, che dal tradizionale rapporto 1:1,33 passò ai due nuovi formati 1:1,66 e 1:1,85 (detti anche panoramici). Dopo diverse oscillazioni, invece, il formato wide screen si attestò sullo standard 1:2,35 (cfr. Vitella 2007).
9 Cfr. Foster 1959 e “American Cinematographer” 1968.
10 Una flessibilità ancora maggiore era garantita dalle pellicole a colori su supporto l6mm invertibile, che divennero funzionali alle esigenze della ripresa televisiva. Alcuni cineasti indipendenti ne fecero il mezzo privilegiato di un’estetica dell’immediatezza, utilizzandolo per effetti di elegante sciatteria in interi film o in sequenze improntate a una differenziazione del registro visivo.
“American Cinematographer” (1968), Eastman Introduces Three Revolutionary New Films, in “American Cinematographer“, XLIX, n. 6, giugno 1968, pp. 424–426, 449 e 466–467.
Foster, Frederick (1959), A Faster Color Negative, in “American Cinematographer“, XL, n. 6, giugno 1959, pp. 364–365 e 368–370.
Lev, Peter (2003), Transforming the Screen. 1950-1959, Charles Scribner’s Sons, New York.
Mazzei, Luca; Vitella, Federico (2007), Geometrie dello sguardo. Contributi allo studio dei formati nel cinema italiano, Carocci, Roma.
Vitella, Federico (2007), Una questione di standard. Il passaggio dall’Academy ai formati panoramici, in Mazzei/Vitella 2007, pp. 65–130.”
(Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 216–218.) (in Italian)
“For our comparisons, original photography was done on 35mm Eastman Color Negative Film, Type 5250, and 16mm Ektachrome Commercial Film, Type 7255. The original films, each edited to one-minute projection time, were printed in a number of ways. Duplication from Eastman Color Negative Film was done using Eastman Color Intermediate Film, Type 5253 or 7253, for the master positive and duplicate negative stages. Prints from Ektachrome Commercial originals were made using Eastman Color Internegative Film, Type 7270. All prints from internegatives were made on Eastman Color Print Film, Type 7383.
Since printers need to be of very high quality in 8mm work, some comments on our equipment are necessary. All contact printing was done on either a 35mm Bell & Howell subtractive printer or a 16mm Bell & Howell additive printer, each equipped with internal air jets. An Acme printer was used for optical printing with these lenses:
1:1 103mm Kodak Printing Ektar at f/4
2:1 90mm Kodak Printing Ektar at f/3.8
4:1 75mm Kodak Enlarging Ektar at f/5.6
Slightly sharper results would have been obtained at 4:1 reduction if the Acme printer could readily be modified to accept the 90mm Ektar lens at the proper conjugates. The 75mm Ektar lens needed only an adapting mount, however, so it was used for these tests. In choosing printer lenses, each lens should be tested for sharpness, and then used only at the optimum f/number. Printers must be in perfect condition to minimize weave, unsteadiness, poor contact, and all factors that hurt image quality and are especially apparent in 8mm work.
It is important that each printing stage be optimized if quality is to be preserved in duplicating. Losses are inherent in each transfer of the image, a fact pointing toward systems with the minimum number of printing operations. As a general rule, image reduction should be postponed to the last stages of printing for best quality; that is, the image should be kept as large as possible as long as possible. Logically, then, the best quality would be expected from a 35mm original that was release printed at 4:1 reduction. In practice however, a duplicate negative would usually be required, and with presently available equipment 4:1 release printing may not be economically feasible.
Since the projected area of the 8mm frame is only 20% of that of a 16mm frame, dirt on 8mm prints is extremely objectionable. The only defense against dirt and other physical defects is careful film handling and cleaning techniques. More than once, a stereo microscope was used to clear the printer gate of an apparently invisible wisp of dirt that had printed through an embarrassing number of times before detection.
Enlargements from several of the 8mm films are shown in Fig. 1. These prints, made using 35mm panchromatic negative, are a poor substitute for viewing the actual color motion pictures, but some understanding of relative sharpness and graininess may be gained by comparing 8mm systems in these reproductions.
A rating of the prints compared in our study is given in Table I. Each system is ranked on both quality and practicality factors, and comments on the obvious features of each system are added.
On the basis of sharpness, 16mm original photography is nearly equal to 35mm original photography when their respective 8mm prints are viewed. A number of cross-comparisons are needed to substantiate this statement, but it should be remembered that one less printing cycle is required in 16mm duplicating than in 35mm duplicating. This, together with the limiting response of the print materials at 8mm image size, tends to offset the size advantage of the 35mm original.
Optical release printing is needed at present for satisfactory screen quality. It would be very desirable to use the faster contact-printing system in making release prints, but most viewers consider an optical print from 16mm a necessity for adequate sharpness. Cartoons or other material not containing fine detail might be suitable for 8mm contact release.
Kodachrome II film has been included in this study because it is a film against which customers can judge their purchases in the commercial market. In picture comparisons the color quality and sharpness of the original film are especially prominent; the long tone scale in the original and the differences in grain quality are also important.
Perhaps our experiences can be summarized in these four conclusions, some of which apply to printing in other than 8mm systems.
(1) Optical release printing from a 16mm or 35mm internegative is required for a satisfactory 8mm print.
(2) In 8mm printing a 16mm original gives nearly the same quality as a 35mm original.
(3) In printing through several stages, film size should be kept as large as possible as long as possible for best quality.
(4) An 8mm Kodachrome II original is sharper and of higher quality than any 8mm print from a 16mm Ektachrome Commercial original or a 35mm Eastman Color Negative original.
A number of printing systems were investigated that have not been mentioned. Shown below as print masters is the use of Ektachrome Reversal Print Film, Type 7386; Ektachrome Commercial; and Eastman Reversal Color Print Film, Type 5269:
Each of these films, in either 8mm or 16mm, has some advantage in grain, color quality, or sharpness, but the overall screen quality of the final print is not equal to that obtained in the Eastman Color system. Furthermore, the 8mm commercial market must be one of high volume, a condition requiring the use of lower-cost print materials.
Kodachrome II Originals
In 8mm prints the emphasis is on sharpness. This has led to many inquiries on the use of 16mm Kodachrome II as an original for release printing. While Kodachrome II is certainly sharper than ECO, it is a higher-contrast film designed for direct projection and not as an original for printing. Prints made from Kodachrome II onto 7270 or 5269 are very high in contrast, except for a low relative red contrast. The contrasts can be made nearly equal by use of special printing filters, but the prints, while sharp, still have high overall contrast.
In addition, our comparisons of prints from Kodachrome II and ECO show considerably less sharpness difference than does a direct comparison of the two originals. The sharpness of Kodachrome II has reached a level where the ability of the printers to transfer the finest detail is a significant factor. Ideally, further improvements in either original or print films should be accompanied by improvements in printing techniques.
After seeing these prints, most viewers agree that for acceptable quality an 8mm print must be prepared from a 16mm or 35mm prior stage in an optical reduction printer. This printing procedure is generally slow and expensive, although there are exceptions in some laboratories. In general, however, it would be advantageous to contact-print from an 8mm internegative, printing two or four rows of 8mm pictures at once. While we cannot yet do this and produce satisfactory quality, examination of contact-printing methods is an active program throughout the industry.
Interest in contact printing is further increased by the question of photographic vs. magnetic soundtracks. If equipment and films would permit four photographic soundtracks to be printed in a single-pass contact printer, considerable reduction in laboratory costs would be realized compared to the present striping and recording operation using magnetic materials. A one-pass picture and sound printer using 35/32mm film with four rows of 8mm pictures is an attractive goal for further development work. Many questions arise, however, when considering photographic sound – the kind of soundtrack, method of recording, equipment for playback, and wear characteristics. These problems, along with many others, must be solved before photographic sound for 8mm can be a practical reality.”
(Keene, G. T.; Clifford, J. D. (1962): Commercial Systems for Making 8mm Prints. In: Journal of the Society of Motion Picture and Television Engineers, 71,6, pp. 447–449.)
Track of the Cat (USA 1954, William A. Wellman)
“But directors were also foregrounding colors that did not provide clues to, or mirror, the psychological states of the character, or even reinforce in any obvious way the tone of a scene. William Wellman’s Track of the Cat provides the extreme example. Although shot in color, Track of the Cat was processed entirely without hue and saturation except for the pinprick of Robert Mitchum’s red woolen coat. The effect is to suggest an obscure allegory, putting into the spotlight an otherwise non-signifying garment worn by a supporting character, the protagonist’s brutish older brother. [Fig 60]. Although still operating within the context of a narrative film, Wellman’s red color spot challenges what Edward Branigan has called “the tyranny of the center of the frame” achieving a certain formal independence highly unusual in Hollywood films – competing for attention with the narrative focal point by wandering off to the margins. The red becomes a non-narrative element that neither enhances Barthes’ ‘reality effect’ nor makes any obvious psychological point. Self-conscious and mannered, Track of the Cat is the closest Hollywood gets, as best I can determine, in opening the gap between color and object.42
42 Edward Branigan, “The Articulation of Color in a Filmic System,” Wide Angle 1/3 (1976), 28–29 shows how Godard transforms this strategy into a full-blown radical system. My analysis of Track of the Cat derives from Branigan’s study.”
(Merritt, Russell (2008): Crying in Color. How Hollywood Coped When Technicolor Died. In: Journal of the National Film and Sound Archive, Australia, 3,2/3, pp. 1–16, on p. 11.)
East of Eden (USA 1955, Elia Kazan)
“Meanwhile, Elia Kazan and his cinematographer Ted McCord devised a highly unorthodox color plot for Kazan’s first color film, East of Eden, built around principles of what may be called “psycho-chromatics,” the arresting idea that some colors, like human behavior, contain opposing and conflicting impulses, depending upon context. For Kazan, the Oedipal director par excellence, green became the master color, an integral part of the psychological texture of the film. [Fig 61]
Green is sometimes considered the great neutral, the color of repose, of pacifism. But it is also connected with envy. It combines associations of growth, with immaturity – the refusal to grow up. [Fig 62]
In our culture it is also the color of money and luck. But above all for Kazan, it contains the associations with both rebirth – freshness – and decay, sickness [“green around the gills”], weakness. And the final scene is keyed to absorb all those contradictions. Whereas the color plots of Technicolor musicals and swashbucklers tended to culminate in festive kaleidoscopic pageantry, East of Eden ends in monochrome – in a room, seen only at the end of the film, dominated by luminous green walls, unevenly lit and crossed with achromatic, graded shadows for an eerie, iridescent effect. Raymond Massey, James Dean’s father, is dying, disillusioned and alone, while Dean, as his son Cal, tries his best to tend to him. [Fig 63]
Kazan called the green room “death’s version of [the old man’s] valley,” the color of vegetation turned into sea-green sickness.43 But the room is also the nacreous site of Cal’s reconciliation with his father, and Abra’s reconciliation with Cal.
With East of Eden we enter a phase of color development in Hollywood film where colors can be indexical, iconic, or simply tied to their symbolic function in a particular culture.
43 Kazan to Michel Ciment in Ciment, Kazan on Kazan (New York: Viking Press, 1974), 123.”
(Merritt, Russell (2008): Crying in Color. How Hollywood Coped When Technicolor Died. In: Journal of the National Film and Sound Archive, Australia, 3,2/3, pp. 1–16, on pp. 11–12.)
Dial M for Murder (USA 1954, Alfred Hitchcock)
“This chapter deals with the construction of cinematic space in Hitchcock’s first experiments with color: Rope (1948), Under Capricorn (1949), and Dial M for Murder (1953). These films are characterized by a rather restrained palette contributing to the creation of a claustrophobic interior, which is central to the plot of these films. Unmistakably in line with the contemporaneous “low key” genres of the Gothic melodrama and film noir, Hitchcock’s earliest color films clearly differ from the glaring colors of his later films that established him as one of the great colorists of film history although only fifteen of his fifty-three films are in color.
In contrast with his films of the later 1950s, in which color is first and foremost invoked to construct exterior public spaces, Hitchcock’s earliest color films use color to create a specific atmosphere that is inherently linked to domestic interiors. In films such as Rope, Under Capricorn, Dial M for Murder, and Rear Window (1954), the narrative is largely or even exclusively situated in domestic spaces that become oppressive and claustrophobic (Jacobs 2007). This tallies with Hitchcock’s Gothic melodramas of the 1940s that present the house as a place of confinement (Doane 1987: 123–154; Waldman 1991). […] By turning the familiar domestic environment into a place of fear, Hitchcock developed into a master of the uncanny. The color films Rope, Under Capricorn, and Dial M for Murder are variations on this Gothic theme, which are set in domestic spaces: a Manhattan penthouse, a Georgian country house in New South Wales, and an apartment in a London Edwardian Mansion House respectively. What’s more, these houses are places where a murder takes place and/or where a character gets the impression that he or she will be killed by another inhabitant. Instead of a place of safety and domestic bliss, the house becomes a trap and an uncanny place of treason, suspicion, and danger. Light and color play an important part in this process.
dial m for murder
Hitchcock’s third color film was Dial M for Murder, which tells another story about somebody trying to kill another resident of his own house. This time, the story is situated in the ground floor apartment of a London mansion. Its interior is marked by sober and elegant classical decorations. The living room has eggshell colored walls and is richly furnished with seats, armchairs, side-tables, and a writing desk. In addition, the apartment is richly decorated and the sophistication of its residents is expressed by the presence of some artworks. A Warner Brothers press release states that “because he is a man of taste and culture, Hitchcock hand-picked many of the props, including an original Rosa Bonheur oil painting, long hidden in Warners’ property gallery, and a pair of valuable Wedgewood vases” (Jacobs 2007: 107). Some of these art works, such as the Bonheur painting, fauve paintings, and a postimpressionist flower piece, render touches of color to the restricted palette of the interior, which is beautifully rendered by Robert Burks’ Warnercolor photography.
Again, the subdued colors emphasize the feeling of an enclosed and confined space. Its boundaries are not only underlined by the motifs of the locked door and the latchkey but also by the closed large window curtain – its dark green called “oppressive” by Johnson (1966: 9–10). In addition, the boundaries of the confined space are marked by a series of high angle shots whereas the claustrophobic feeling is further elaborated by the 3-D process. Although Hitchcock refused to capitulate to the outrageous eccentricities of 3-D gimmickry (no objects are thrown in the face of the viewer), he favored compositions that place lamps, knickknacks, and pieces of furniture of diverging colors in the foreground, walls well in the background, and human subjects in a middle-ground. In addition, a subtle depth of field is created by the camera, favoring low positions and gliding fluently around the furniture. In so doing, Hitchcock used the 3-D process to create dramatic effects, emphasizing the ways in which the apartment serves as a trap for the characters.
Again, the contrast between light and darkness contributes to this effect. Crucial scenes such as the planning of the murder, the murder itself (actually an attempted murder that turns into the killing of the murderer), and the exposing of the criminal are shown in darkness – with the light of the fireplace, the light from an adjacent room coming under the door, or a flash light as dramatic elements. Against the quiet and almost monochrome background of harmonious colors of the interior, the changing colors of Margot’s dresses are remarkable. Wearing a color perfectly in harmony with the background in the company of her husband in the opening scene, Margot (Grace Kelly) is dressed in sparkling red in the subsequent scene with her American lover. As Laurent Fiévet (1995: 72–73) has noted, Hitchcockian red is often associated with the red theater curtain. Since Hitchcock heroines tend to appear late in the film, they are often surrounded by red indicating that the plot truly begins. According to Fiévet, in Hitchcock, red is also the color of fake appearances and of lies and he mentions Margot as an example. Later in the film, she is shown with a blue night gown during the murder scene as if she has merged into her domestic setting. The day after the murder, she is dressed in dark blue and she wears a brown coat in the final scene blending with the background again. Like Milly, the housemaid in Under Capricorn, she repeatedly becomes part of the interior. Referring to Scottie in the hospital room in Vertigo, Marnie disappearing in the urban gray, and Roger Thornhill in the vast wasteland of North by Northwest, Allen (2006: 134) noted that “loss of identity or emotional vacuity is evoked in precise moments in Hitchcock films through the loss of color.” Margot is also wearing dark clothes during the scene of the trial, which is in many ways remarkable in respect to color. Apart from some establishing shots of the apartment’s exterior, a shot in a cab, a view of an arriving ship, and some shots in a restaurant, the courtroom scene is the only scene that does not take place in the apartment itself. It is a scene, however, without spatial coordinates or in which space is exclusively constructed by color and light. Only a medium close-up of Margot and the judge are visible and they are shown against a monochrome background, which changes from blue to red. It is as if this red reverberates the blood that intruded and destroyed the home, which in Hitchcock is never a site of domestic bliss. Years before Marnie’s obsession with red, the blood stain on the carpet in this bourgeois apartment acts as an intriguingly strange object. It is not only a literal stain on the floor but also an uncanny detail in the mise-en scène, a stain in the visual field or what Zizek (1999) has called a “Hitchcockian blot.”
the restrained mode
In his three first color films (and, to a large extent, also in his next color film Rear Window), Hitchcock favored muted colors before he started to celebrate the exuberance of glorious Technicolor from the mid-1950s onwards. This line of approach was already announced by Hitchcock in 1937, ten years before his first color film:
I should never want to fill the screen with color: it ought to be used economically – to put new words into the screen’s visual language when there’s a need for them. You could start to color film with a boardroom scene: somber paneling and furniture, the director’s all in dark clothes and white collars. Then the chairman’s wife comes in wearing a red hat. She takes the attention of the audience at once, just because of that one note of color. (Hitchcock 1937: 258)
In so doing, Hitchcock’s early color experiments tally with what Higgins (2007: 76–136) terms the subdued or restrained mode of Technicolor filming which flourished in the late 1930s. […] Hitchcock’s first color films are characterized by narrow palettes and tight harmonies although this restrained mode allows for an occasionally overt display of color, dramatic punctuation, and elegant decoration. This predilection for muted colors can also be linked to the generic conventions of the somber domestic melodrama and the crime thriller, which both in the 1940s and early 1950s reached their apex in the “black-and-white” and “low key” genres of the Gothic romance film and film noir. Together with John Stahl’s Leave Her to Heaven (1946), Rope was one of the first crime thrillers shot in color. In addition, Hitchcock’s subdued palette contributes to the creation of a claustrophobic interior, which is central to the plot of these films. Consequently, color plays an important part in the integration of the characters in the surrounding interior. However, blending the characters with their environment entailed specific problems in color photography. On the set of his first color feature Rope, Hitchcock was surprised to see the extent to which actors were separated from the background through the use of backlights even though in color there was no need for this (unless the actor should be dressed in exactly the same color as the background). In his famous interview with François Truffaut (1984: 183), Hitchcock stated that he truly believed that “the problem of lighting in color film has not been solved. […]” Last but not least, in Rope, Under Capricorn, and Dial M for Murder, specific colors, color combinations, and pictorial light effects help to mark the difference between inside and outside, and hence often between on screen and off screen. In so doing, colors greatly add to the construction of the motif of intrusion.
Allen, R. (2006) “Hitchcock’s Color Designs”, in A. Dalle Vacche and B. Price (eds) Color: The Film Reader, New York: Routledge: 131–44.
Doane, M. A. (1987) The Desire to Desire: The Woman’s Film of the 1940s, Bloomington and Indianapolis, IN: Indiana University Press.
Fiévet, L. (1995) “Vertiges chromatiques, Alfred Hitchcock: Vertigo“, in J. Aumont (ed.) La Couleur en cinéma, Paris: Cinémathèque française: 72–73.
Higgins, S. (2007) Harnessing the Technicolor Rainbow: Color Design in the 1930s, Austin, TX: University of Texas Press.
Hitchcock, A. (1937) “Direction”, in S. Gottlieb (ed.) (1995) Hitchcock on Hitchcock: Selected Writings and Interviews, London: Faber & Faber: 253–261.
Jacobs, S. (2007) The Wrong House: The Architecture of Alfred Hitchcock, Rotterdam: 010 Publishers.
Johnson, W. (1966) “Coming to Terms with Color”, Film Quarterly 20,1 (Autumn): 2–22.
Truffaut, F. (1984) Hitchcock/Truffaut, New York: Simon & Schuster.
Waldman, D. (1991) “Architectural Metaphor in the Gothic Romance Film”, Iris 12: 55–70.
Zizek, S. (1999) “The Hitchcockian Blot”, in R. Allen and S. I. Gonzalès (eds.) Alfred Hitchcock: Centenary Essays, London: BFI: 123–140.”
(Jacobs, Steven (2013): Color and Containment. Domestic Spaces and Restrained Palettes in Hitchcock’s First Color Films. In: Simon Brown, Sarah Street and Liz Watkins (eds.): Color and the Moving Image. History, Theory, Aesthetics, Archive. New York, London: Routledge, pp. 179–188, on pp. 179–187.)
Lola Montès (FRA/BRD 1955, Max Ophüls)
“La couleur enveloppe le décor au point d’en brouiller les repères, d’en gommer les contours. Elle ne signale rien mais au contraire efface les traces de manière que seule sa présence fasse certitude.
La chorégraphie ne dispose pas du corps de Lola, seulement des couleurs qui l’entourent, et de ces couleurs jaillissent formes et mouvements, qui agencent, coordonnent et donnent l’illusion, presque parfaite, que ballet il y a. Que les couleurs provoquent du mouvement, la peinture l’a souvent démontré2, mais ici les couleurs régissent le déroulement du film, le contraignant à s’agencer selon un rythme qui est le leur. Pour mouvoir la danseuse sous le chapiteau, les couleurs se font tour à tour sélectives, aériennes et musicales.
Sélectives: restreintes dans l’image, les couleurs restituent la sensation qu’elles provoquent en peinture. Cet effet, très fréquent dans les ballets des comédies musicales, consiste, à partir de strates de couleurs, à créer un nivellement à l’intérieur du plan, ainsi accusé dans sa profondeur. Les bi- ou tri-chromies transforment l’espace, qui semble s’ouvrir et se fermer au gré des gestes et figures des danseurs.
On relève principalement dans Lola Montés deux séries de couleurs:
– une opposition à trois termes, noir/rouge/blanc, évoquant de près ou de loin la danse. Ce sont les smokings et les hauts-de-forme bicolores des “hommes-monnaie”, qui prennent en charge la descente d’escalier faussement majestueuse de Lola, en absorbant son tutu lamé, lui aussi noir et blanc. Quelques plaques rouges, posées sur les sacs à main et les rideaux qui pendent de chaque côté du cadre, renforcent la symétrie provoquée par ce groupe qui coupe le plan en deux, pour mieux laisser régner Lola en son centre. Encore plus probant, le rouge, porté par la galipette d’un acrobate, qui s’agite devant le tableau d’Aracuse en ombres chinoises. Couleur-mouvement, le rouge s’oppose ici à la fixité de l’image en noir-et-blanc. Un rouge pour annoncer la mise en marche d’une image passée, le déclenchement d’un appareil, grâce à qui l’image soudain s’animera?
– l’orange et le bleu se déclinent différemment selon les séquences des flashes-back qui présentent toute une gamme de bleus, du bleu dur au bleu gris, et d’orange, assimilés le plus souvent aux couleurs de la terre (brun pour la rupture des époux, cuivre et ocre pour l’épisode avec Liszt). Ces couleurs se retrouvent sous le chapiteau, où la bichromie prend valeur d’opposition (elles sont complémentaires). À une opposition formelle entre le haut et le bas, répondent plusieurs contrastes relevant de ce qu’Ophuls considérait comme le “rôle psychologique des couleurs dans le spectacle, de l’expressivité dramatique des couleurs”3: contraste entre le sommet bleuté de la tente, vers lequel se hisse Lola, et le sol orangé duquel elle émerge4, entre les cieux, les ténèbres et la terre, le sable dans lequel elle s’enlise, ou la couleur du fauve auquel la danseuse est sans cesse comparée, contraste entre le froid (la fraîcheur de l’altitude) et le chaud (la chaleur de la terre), mais aussi l’envol et la chute.
Aériennes: les couleurs ont-elles un poids? Sont-elles lourdes, sont-elles légères?
Le jaune, absent des souvenirs, prend au cours du spectacle un statut particulier. Doré, il s’anime. Pour contrecarrer l’immobilité des jongleuses – autant de doubles de Lola, aux visages livides cerclés de noir – le jaune, déposé sur des quilles, s’allège, s’élève et parsème le cadre de touches étincelantes. De même, la couronne qui traverse le cadre d’un mouvement vertical, creuse un sillon doré et participe à la scénographie de l’ensemble. Ainsi suspendue, la couleur échappe à la pesanteur et voltige à la place des corps, vissés au sol.
Cette envolée colorée contraste avec l’intervention des faisceaux orange et bleu, qui ne cohabitent plus dans le cadre mais alternent dans la durée, si bien que la mobilité de la couleur n’est en aucun cas justifiée par un corps ou un objet. Les couleurs se détachent de leurs supports. Couleurs-matières, elles deviennent couleurs-lumières et, libérées des corps, se diffusent dans l’espace pour recouvrir les autres couleurs d’un faisceau monochrome. Mais, dotée d’un tel pouvoir, totalisant et totalitaire, la couleur, écrasante, semble s’affaisser d’avoir effacé les couleurs du monde; l’image s’aplatit, l’espace se resserre, alourdi par son uniformité. La couleur retombe à terre.
Musicales: du mouvement provoqué par une sélection de couleurs à la musique, il n’y a qu’un pas, car la couleur, en se déplaçant, crée un rythme qui lui est propre.
Chacun des faisceaux colorés est précédé d’un roulement de tambour, lui-même accompagné d’un plan sur l’imposant lustre qui pend sous la grande toile du chapiteau, comme si la musique, au même titre que la couleur, émergeait de cette source lumineuse. Parce qu’ils emportent les couleurs sur leur passage, les faisceaux lumineux s’apparentent à la tonalité musicale5.
Lors de la scène de la chanson, des faisceaux orange et bleus rythment le décalage des pistes sonores. En superposant son chant à celui d’une voix enregistrée, l’écuyer brouille les pistes sonores si bien que l’on ne distingue plus réellement de qui émane la voix qui chante. Activées par les couleurs, les deux voix s’élèvent et conduisent l’écuyer jusqu’au haut des marches.
Inversement, transformée en sirène, Lola feint de nager derrière des cadres rectangulaires, bandes achromes qui effacent jusqu’aux contours de son corps. La décoloration de cette traversée aquatique, projetant la danseuse dans l’image délavée d’une vieille carte postale – fantaisie ou image d’Épinal – va de pair avec les vers de mirliton de l’écuyer (“En Turquie, sur la demande du Sultan / elle se baigne presque nue, c’était passionnant / les invités furent contents, les poissons également …”). L’eau absorbe les couleurs, qui, effacées, ralentissent les gestes de la danseuse, et retient la musique. Les rimes et assonances qui racontent, à la manière d’une chanson de geste dérisoire, les exploits de Lola, insistant sur leur aspect picaresque, n’accèdent pas au chant, comme si l’achromie empêchait la musique de s’épanouir.
“Elle dansait – silence enveloppé de lumière / Brume dans une clarté de lune / Une musique qui charmait la vue / Mais à l’oreille étrangère.”6
La chorégraphie mise en œuvre sous le chapiteau n’est pas le résultat d’une instrumentation musicale mais celui d’une composition de couleurs. La couleur fait danser Lola, ce corps dont l’image s’effacerait sans doute sans les débordements colorés qui l’empêchent de sombrer tout à fait; car la couleur aspire le corps. Ce sont encore les faisceaux bleus et orange, dont la luminosité recouvre le corps de Lola au point de le transformer en ombre.
Emportée sur le cheval du diplomate, Lola voit son corps enlevé par un mouvement qui n’est pas sien. Le diplomate cache Lola dans son palais: seule une ombre colorée transparaît de l’épais rideau derrière lequel le couple s’embrasse. La couleur capte le corps de la danseuse; dénaturée, l’image est défigurée. Seule figuration possible pour un corps qui a perdu le contrôle de ses gestes, corps cadavérique qui gît dans l’arène d’un cirque: une ombre à la merci de la couleur, un corps dépourvu de matière au profit de la lumière qui l’éclaire.
De ce corps, plus tard, il ne restera qu’un nom. Vainqueur d’une bataille d’ombres chinoises, le garde du corps Bugakof troue la toile pour faire jaillir d’un écran noir et blanc les marques colorées des capitales où la danseuse a fait scandale. Inscrits à même la peau, les noms des villes sont traversés d’une large écharpe portant le nom de Lola. Le corps tatoué de bleu et d’orange s’est approprié les couleurs maîtresses de Lola Montés. Pour posséder le corps de la femme dont il est amoureux, le garde du corps s’arrache ses couleurs. Ce “garde des couleurs” dépossède le corps de Lola de son propre nom.
Le roi fait peindre Lola: “Un jour, il lui demande si elle ose apparaître pour lui tout en rose, elle ose. Et le roi, épris par sa pose, lui offre un palais. C’est déjà quelque chose”, explique l’écuyer. L’assonance immobilise encore une fois la danseuse et le rose la dénude. Pire, la couleur dépouille le corps, l’incarcère dans une toile et le fige en une image. Le rose dit la nudité; la couleur fait corps avec l’image. Comme dans le Nu rose de Matisse, le corps est couleur et rien d’autre.
La couleur mutile le corps, transformant la poupée de porcelaine en ombre évanescente; après son identité et sa peau, son corps est aussi déchiré dans sa chair. Quelle image possible pour ce corps décoloré, évidé? Sa propre représentation ne va plus de soi: au palais de Bavière, le roi ne trouve pas de murs pour accrocher sa toile.
Du corps, rien ne subsiste. Le visage, avec difficulté tente de résister.
Au début du spectacle, des hommes déposent devant Lola, juchée sur un piédestal, une couronne garnie de têtes à son image. Bientôt un autre groupe succède à ce premier cortège et détache les longues tiges au bout desquelles pendent les visages de céramique. À l’instar des bourreaux qui, pendant la Révolution, exhibaient devant la foule les têtes des guillotinés, ces hommes, tout de rouge vêtus, arborent fièrement devant eux les visages de la jeune femme, tandis que les roulements de tambours, qui accompagnent les allers et venues des personnages, rappellent la musique martiale des mises à mort.
Mais de quoi la danseuse est-elle coupable? Au XIXe siècle fleurissent les traités – vaguement inspirés des théories de Darwin – qui prétendent déceler la nature de la gent féminine, en comparant notamment la criminelle et la prostituée. L’une a la tête tranchée, l’autre est corps et rien d’autre7. Lola la courtisane doit faire don de son visage à ses bourreaux, eux-mêmes destitués de leurs figures par la couleur rouge qui s’élève jusqu’au haut de leurs crânes, effaçant les reliefs des traits. La couleur accuse le corps, elle le touche dans sa matière, l’atteint dans sa forme. Lola en perd la tête, qui accompagne, le temps d’une représentation, le corps des hommes, c’est-à-dire le corps de l’Autre par excellence. Elle est aliénée, étrangère à elle-même. En prostituant son corps, Lola a aussi vendu son âme.
Le premier souvenir est introduit par un long fondu, au cours duquel le buste de Lola, coloré d’un faisceau orange, traverse une forêt, glisse le long d’un chemin, remonte le cours d’une rivière avec une grâce aérienne, simulée par le bleu des arbres et de l’eau. Les couleurs du chapiteau résistent à l’envahissement des éléments naturels au point de se greffer sur l’image du passé. La couleur de la robe s’impose; lentement, elle s’immisce dans le véhicule. L’ombre colorée du corps ne s’effacera qu’une fois superposée à la roulotte: c’est la superposition de cette couleur sur elle-même qui clôt la surimpression. Puis, lentement, la caméra se rapproche de la petite fenêtre derrière laquelle Lola, étendue, fume. La caméra se fige alors de manière à placer le montant horizontal de la fenêtre sur l’encolure de la robe. Cette barre orange atteste de la vaine résistance des couleurs noire et blanche de l’habit de Lola face au débordement des teintes cuivrées qui ont envahi la berline. La couleur tranche le corps et lui coupe la tête.
À tête tranchée, la parole est coupée. Lola n’a plus son mot à dire. Sans cesse interrompue, elle n’est plus maître de ses souvenirs, investis par le récit de l’écuyer, si bien que la représentation déteint sur la réalité. “L’image actuelle et l’image virtuelle coexistent, elles entrent dans un circuit qui nous ramène constamment de l’une à l’autre, elles forment une seule et même “scène” où les personnages appartiennent au réel et pourtant jouent un rôle. Bref, c’est tout le réel, la vie toute entière, qui est devenue spectacle.”8 C’est l’invasion de la couleur orange dans la roulotte, qui s’étale grossièrement sur les moindres objets. Orange du chapiteau qui, dans sa vulgaire uniformité, enferme et étouffe la danseuse.
La violence du couperet est aussi radicale que l’opposition entre le noir et le blanc. La blancheur immaculée de la tenue de mariée du spectacle disparaît dans le souvenir sous le revers d’une robe noire de maison.
Mais, plus que tout autre, un tranquille glissement de regard signale le basculement intime dans la couleur, lors de la première rencontre de Lola et de l’écuyer. Du coup d’œil jeté au paysage blanchi par la neige, à l’escalier que l’on devine derrière les vitraux orange et bleus de la porte d’entrée, le regard qui surveille l’arrivée de l’écuyer est tronqué. Lola perd la parole et son regard, touché lui aussi par la couleur, ne distingue plus les couleurs du monde.
La chute ultime, une dernière fois unira le bleu à l’orange. Avant de sauter, Lola, éreintée, baisse ses paupières maquillées d’un filet de bleu pailleté. Le bleu aveuglant s’est incrusté dans le corps de la danseuse, à la source du regard. “Glissant vers le noir [le bleu] prend la consonance d’une tristesse inhumaine. Il devient un approfondissement infini dans des états graves qui n’ont pas de fin et qui ne peuvent en avoir.”9
Oppressée, exsangue, Lola abandonne le bleu et laisse tomber son corps jusqu’à la cage. Libératrice, sa chute procure aussi la jouissance de sentir, enfin, son corps chavirer dans les airs et échapper momentanément à l’emprise de la couleur. “On s’aperçoit que le sens même de la danse est sa chute et que ce sens demeure caché dans l’abandon du corps à sa pesanteur, dans l’oubli de soi en faveur d’un espace sans orientation […]. La chute entraîne le corps dans le renversement et la mine d’un monde perçu en descendance, pour le faire retrouver à l’instant, sous l’effet de son poids, la douleur même comme réalité corporelle.”10
La chute d’une couleur à l’autre marque l’aboutissement pour un soir du trajet de Lola vers la douleur. La douleur du bleu, Lola la connaît bien, qui, encore jeune fille, la rencontre à bord d’un bateau la conduisant vers l’Europe. Ce soir là, Lola accomplit son parcours initiatique, celui qui va de l’orangé au bleu et qu’elle répétera sa vie durant. Sur le bateau, une porte se referme derrière elle, laissant transparaître les grilles brunâtres d’une cage à venir. Lola longe le couloir puis s’arrête devant le cadre bleuté renvoyant l’image d’un spectacle: la mère qui valse avec son amant. La ressemblance entre les deux femmes est frappante. En se superposant, leurs robes, taillées dans le même tissu bleu clair présagent le mariage entre Lola et le prétendant de sa mère. Lors de ce clair-obscur coloré, Lola assiste au déroulement de sa vie. L’image de son corps s’évanouit et elle laisse échapper sa peine.
Au bout du couloir, le pont. Sous le bleu velouté de la nuit, Lola lève les yeux vers le ciel et la lueur qui éclaire un instant son visage révèle ses larmes. Les lumières de la nuit s’éteignent et le décor s’abaisse, qui dévoile un ciel de pacotille, d’un bleu beaucoup moins pur: le rideau d’une représentation qui s’achève.
2 “Le statique ne fait pas obstacle au sentiment du mouvement. C’est un mouvement placé à un degré d’élévation qui n’entraîne pas les muscles des spectateurs, mais seulement leur esprit” (Henri Matisse, Écrits et propos sur l’art, Hermann, 1972, p. 63).
3 Ophuls ajoute: “Les unités de couleurs ne sont pas des unités de décor à décor, ce sont des unités d’atmosphère dramatique” (propos rapportés par Georges Annenkov, Max Ophuls, Le Terrain Vague, 1962, pp. 88–89).
4 Pour Kandinsky, le bleu est la couleur typiquement céleste, le jaune, la couleur typiquement terrestre.
5 “Ce qu’on appelait jusqu’ici le ton était le voile d’une coloration unique jetée sur l’ensemble du tableau. […] Ce ne serait pas faire erreur que de comparer un tableau d’un effet intense à un morceau de musique en majeur, et un autre d’un effet adouci à un morceau en mineur” (Johann Wolfgang Goethe, Traité des couleurs, Triades, 1973, p. 256).
6 Dylan Thomas, Vision et prière et autres poèmes, Poésie/Gallimard, pp. 101–102.
7 La représentation de la prostituée comme une femme sans tête est très fréquente dans les dessins de Magritte ou de Labisse.
8 Gilles Deleuze, L’image-temps, Éd. de Minuit, 1985, p. 112.
9 Wassily Kandinsky, Du Spirituel dans l’art et dans la peinture en particulier, Denoël-Gonthier, coll. “Médiations”, 1969–1976, pp. 148–149.
10 “La danse – entre fondation et ruine”, in Doina Maria Petresco, Danse et pensée. Une autre scène pour la danse, Sammeron, Germs, 1993, p. 306.”
(André, Emanuelle (1995): La couleur douloureuse. Max Ophuls. Lola Montès. In: Jacques Aumont (ed.): La Couleur en cinéma. Mailand: Mazzotta, pp. 120–122.) (in French)
“In a three-color process, the accuracy of reproduction is greatly increased and the freedom of choice is greatly restricted” (Ball 1935), but gives no further details. Similarly Henry Willhelm, doyen of photographic conservators, devotes a chapter to “The Extraordinarily Stable Technicolor Dye-Imbibition Process” in which he remarks that “the dyes in the Technicolor imbibition process have better spectral characteristics than dyes available for current negative-positive processes” (Willhelm 1993, 355). Technicolor prints that have had a tough road life can show some slight fading of the cyan dye, for example, but Eastmancolor’s cyans fade radically and quickly.
Ball, J. A. 1935. The Technicolor Process of Three-Color Cinematography. Journal of Motion Picture Engineers XXV (2): 127–138. http://www.widescreenmuseum.com/oldcolor/ball.htm.
Wilhelm, Henry. 1993. The Permanence and Care of Color Photographs: Traditional and Digital Color Prints, Color Negatives, Slides and Motion Pictures. Grinnell, IA: Preservation Publishing Company.”
(Cubitt, Sean (2014): The Practice of Light. A Genealogy of Technologies from Prints to Pixels. Cambridge, Massachusetts, London: MIT Press, on p. 135.)
Anonymous (1980): Colour Problem. In: Sight and Sound, 50, pp. 12–13, on p. 12.
O’Connell, Bill (1979): Fade Out. In: Film Comment, 15,5, pp. 11-18.
Scorsese, Martin (1980): Letter. In: Film Comment, 16,1, 1980, p. 79.
Brown, Simon; Street, Sarah; Watkins, Liz (eds.) (2013): Interview. Dave Davis. In: British Colour Cinema. Practices and Theories. Hampshire: Palgrave Macmillan, pp. 123–130, on pp. 125–127.
Dupar, Edwin B. (1952): Warnercolor. Newest of Color Film Process. In: American Cinematographer, 33,9, pp. 384–385.
Deux ou trois choses que je sais d’elle (FRA 1966, Jean-Luc Godard):
Branigan, Edward (1976): The Articulation of Color in a Filmic System. Deux ou trois choses que je sais d’elle. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 170–182, on pp. 170–180.
Dial M for Murder (USA 1954, Alfred Hitchcock):
Jacobs, Steven (2013): Color and Containment. Domestic Spaces and Restrained Palettes in Hitchcock’s First Color Films. In: Simon Brown, Sarah Street and Liz Watkins (eds.): Color and the Moving Image. History, Theory, Aesthetics, Archive. New York, London: Routledge, pp. 179–188, on pp. 179–187.
East of Eden (USA 1955, Elia Kazan):
Merritt, Russell (2008): Crying in Color. How Hollywood Coped When Technicolor Died. In: Journal of the National Film and Sound Archive, Australia, 3,2/3, pp. 1–16, on pp. 11–12.
Lola Montès (FRA/BRD 1955, Max Ophüls):
André, Emanuelle (1995): La couleur douloureuse. Max Ophuls. Lola Montès. In: Jacques Aumont (ed.): La Couleur en cinéma. Mailand: Mazzotta, pp. 120–122. (in French)
Lust for Life (USA 1956, Vincente Minnelli):
Jacobs, Lewis (1970): The Mobility of Color. In: Lewis Jacobs (ed.): The Movies as Medium. New York: Farrar, Straus and Giroux, pp. 189–196, on p. 193.
Mon oncle (FRA/ITA 1958, Jacques Tati):
Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 223–224. (in Italian)
Pierrot le fou (FRA 1965, Jean-Luc Godard):
Branigan, Edward (1976): The Articulation of Color in a Filmic System. Deux ou trois choses que je sais d’elle. In: Angela Dalle Vacche and Brian Price (eds.): Color. The Film Reader. New York: Routledge, 2006, pp. 170–182, on p. 177.
Racconti romani (ITA 1955, Gianni Franciolini):
Pierotti, Federico (2012): La seduzione dello spettro. Storia e cultura del colore nel cinema. Genova: Le Mani-Microart, on pp. 210–212. (in Italian)
Track of the Cat (USA 1954, William A. Wellman):
Bourget, Jean-Loup (1995): Esthétiques du Technicolor. In: Jacques Aumont (ed.): La couleur au cinéma. Mailand: Mazzotta, pp. 110–119, on pp. 114–115.) (in French)
Merritt, Russell (2008): Crying in Color. How Hollywood Coped When Technicolor Died. In: Journal of the National Film and Sound Archive, Australia, 3,2/3, pp. 1–16, on p. 11.
Une femme est une femme (FRA/ITA 1961, Jean-Luc Godard):
Sharits, Paul (1966): Red, Blue, Godard. In: Film Quarterly, 19,4, 1966, pp. 24-29, on pp. 24-26.