Technical Notes
The Dye Transfer Process: Introduction
The Kodak Dye Transfer Process first came on the market in 1945, and it was officially discontinued in 1993. When it was taken out of production I purchased enough of the materials to continue for many years. It may be revived one day in a slightly different form, but for the moment is gone. Therefore, in writing about Dye Transfer I have had some difficulty with my tenses. Because I still use it, the past tense is inappropriate many times, but many things about the process are clearly past.
Dye transfer is a photographic process which is capable of making prints of the highest quality, but most of the tons of Kodak Dye Transfer materials were for advertising and commercial work. When financially possible, artists did often have their photographs printed with Dye Transfer by commercial labs, but only a few printed their own work. This is unfortunate, for with this process it is possible to exercise an extraordinary degree of control over the result. However, it requires considerable time, materials, and patience.
A dye transfer print is more permanent than prints made by conventional processes. When one has color prints made at a local store, the Type-C prints that are produced have many layers. These are very complex chemically and they will fade noticeably in a few decades in dark storage at normal temperatures. When Type-C prints are made, it is possible to add or subtract from either of three basic colors (cyan, magenta, and yellow), and the overall contrast can be altered, but normlly that is about the limit of the controls that are available. With dye transfer, the print itself has no delicate light sensitive emulsions, and it does undergo complex chemical processing. The dye transfer paper is plain except for a mordant and a gelatin layer, and the dyes are relatively stable fabric dyes. Furthermore, long before the print itself is made, the three basic colors are taken apart in such a way that the contrast of each one can be controlled individually; the highlights, mid-tones and shadows can be controlled separately for each color. The black and white color separations that are generally made in order to produce a Dye Transfer print are in themselves an archival record of the image.
It should be noted here that exercising control over the final image is not "cheating". Photographs are rarely an exact reproduction of reality. Technology may now come close to reproducing on a CD sounds exactly as they occurred, but visual reproduction technologies face different problems. Lenses affect color balance, and color films differ considerably in their rendering of any scene from type to type, even from emulsion to emulsion. The human eye is a remarkable instrument in its ability to make relative discriminations, and it is far more sensitive to errors in color tones than to variations in black-and-white density, but it is not normally calibrated to any absolute. As long as flesh tones are reasonable and as long as colors are balanced in relation to each-other, people tend to accept pictures with fairly severe color distortions. Films do not reproduce all colors equally well. Taking a piece of film and printing it literally, by any photo process, is by no means an "accurate" reflection of reality. It is merely a record of how one set of chemicals have reacted to light.
The human eye can see a far wider range of illumination levels than can be rendered by a piece of film, and film can hold a wider range than paper. Outside on a sunny day if we measure the light level in the darkest shadow in which we can still see detail and double it twelve times (twelve stops on a camera, log 3.6), we will have approximately the light level of the brightest highlight in which we can still see detail. From the darkest to the lightest spot on a print there will be only seven to eight such doublings or stops (log 2.3). If a camera records the various light levels of the twelve-stop scene on a film which spans nine stops, and they are printed straight onto paper which can hold seven stops of light, this print will not be an exact reproduction of nature. The shadows will have turned black and the highlights will be burned out.
If a film's sensitivity is plotted on a graph with a log scale, each increase in light produces a uniform increase in silver density on the film. This produces a straight line. At the beginning and end of the film's range of sensitivity the line curves off. There is still useful information on those curved portions of the film's sensitivity, and one of the tasks of masking is to extend the straight line by accentuating the information on those curves. We cannot exactly reproduce the same range of light levels which we see in nature, but we can avoid losing some of the exciting highlight and shadow details. Painters have traditionally flattened the mid-tones, used color to separate objects instead of contrast, then used highlights and black sparingly for accent.
A color photograph is a transformation; a wide range of light levels and a full spectrum of colors are transferred to a medium of finite scope with only those colors which can be rendered by a few chemical dyes. Every print is an interpretation, the result of many choices and compromises, which are either made consciously or by default. With dye transfer it is possible to make these changes with a greater degree of control than with other processes. Films taken in remote places, subjected to extremes of temperature and humidity and kept for many weeks before development are not ideal records either, and dye transfer is useful in restoring these fragile remnants to life. With dye transfer prints it is possible to exercise considerable control over every aspect of the color and contrast, but one must make the process work through trial and error and many small adjustments. Prints made mechanically with a conventional process are to dye transfer prints as a piece of music played on a player piano is to one played on a concert grand.
Concerning my prints, I am often asked by people from whom I would normally expect to hear intelligent questions: "Did it really look like that?" Even if the mechanical systems could reproduce all of the range of light which the eye saw at a given instant, would that still satisfy the memory? A few moments are spent on the other side of the world in a land still feudal and largely nomadic. Goats move about nearby: a donkey brays in the distance. The light is clear at six thousand feet on a brisk spring morning as it breaks through the clouds and flickers across hills outlined in lavender. Across these bright green, rolling hills are patches of red flowers. The landscape is bare of trees, and one can see far across the valley to the dark mountains beyond. Tall grass rustles in the wind. Finally, the man for whom one has waited is in a suitable place; a photo is taken. Did the scene really look exactly like the final print? That is not possible, but one can try to make an object which testifies to the wonder of the original scene and one which is beautiful in its own right. "Did it really look like that?" Does the 1812 Overture sound exactly like the battle?