Color photography is based on the two topics we have already discussed: forming black and white images on film and matching any given color using the three primary colors. In essence, color photography consists of forming three independent images using the light reflected from the object in the 3 primary colors and then combining those 3 independent images to reconstruct the initial color of the object. Each one of the three images is formed using techniques very similar to those that are used for black and white photography as described in the previous topic. The technical tricks involve how the 3 separate red, blue and green images are produced initially and how they are recombined in the final print. Since all color photography is based on using the three primary colors (either by addition or subtraction), it is affected by the same limitations as other processes that use this technique. That is, it is impossible to reproduce colors that are very saturated. Different brands of film use different dyes and slightly different developers, but all of them are basically the same and are affected by this limitation.
As with black and white photography, color photography is divided into two broad categories – (1) color-print films, which produce a negative image after one development and a positive image after a second exposure in the development process, and (2) reversal films, which produce a positive image on the film itself after a more complicated development process. Reversal films are used for slides and movies.
The early color films were usually reversal films for slides or movies, and were based on forming colors by addition of the 3 primaries: red, blue and green. The light from the object being photographed was divided into 3 beams using red, blue and green filters, and the 3 beams were used to produce 3 images. One of the early techniques for achieving this separation used standard black and white film with a distribution of tiny color filters spread more or less uniformly over the entire film area. (The filters were added to the emulsion more or less randomly across the negative during the manufacturing process.) Each filter passed light only in one of the three primary additive colors to the film below. The density of silver below each filter corresponded to the intensity of light in that primary color at that point. The film was then developed using the standard reversal technique, which produces a clear spot on the film wherever the initial exposure contained lots of free silver and a dark spot on the film wherever the initial exposure contained no free silver. When the film was viewed through white light (in a slide projector, for example), the combination of the transparency of the film (which was proportional to the intensity of light from the object at that point) and the tiny filters in front of each point on the film re-created a full color image by the usual method of 3-color addition. (For example, a clear spot underneath a red filter implied that the original object had a lot of red light at that point. When the developed film was projected, the red filter in front of that clear spot passed red light to the screen. Conversely, a dark spot on the film meant that the original object had no red light at that point, because that dark spot was created during the second development. Thus no red light was transmitted at that point. Each of the other filters worked the same way.)
A variation of this process used three physically distinct films, each of which was exposed to the light from the object in only one of the primary colors. Each of the films was then processed to produce a reversal image, and the three films were then projected onto a screen using three independent projectors, each of which was equipped with the proper filter, which only allowed light of one primary color to reach the screen. This method is pretty cumbersome for films, but is exactly the method used in some projection-TV systems.
The primary advantage of systems based on the additive primary colors is that the film itself is standard black and white film, and can be processed by the usual developers to produce a reversal image. This was a significant advantage in the early days of photography, when most developing was done by hand, because color developers are tricky to use and require very careful control of the development time and temperature.
One of the limitations of this class of methods was that it was difficult to make copies of the picture without degrading its sharpness. This degradation resulted from the fact that since the tiny filters were arranged more or less at random across the film surface, different pieces of film had the 3 filters arranged in different semi-random ways. A point that was bright red on one negative might not be copied properly because the film used for the copy did not happen to have a red filter at exactly the same point. The fact that the images in the 3 primary colors are separated in space to some extent is a fundamental limitation of all of these methods, since the resolution of the image is limited by this spatial separation in a fundamental way. (Current TV pictures suffer from the same limitation.) The obvious solution to this problem is to use a process based on subtractive primaries, since the three images can be exactly coincident if subtractive primaries are used.
Most newer films differ in two important respects from the earlier films based on the additive primaries. First of all, they work by color subtraction rather than by addition, and second of all, the different colors are generated using dyes – transparent materials (similar to the inks used for printing) that function by subtraction by absorbing one of the positive primaries. As we mentioned in the previous paragraph, color subtraction is capable of higher resolution in principle, since the three color images can be superimposed on each other to form the final full-color image.
The earliest color films based on subtractive primaries used three physically different pieces of film for each of the three colors. Each film recorded the image in one of the primary colors. Each of the three films was processed so that the thickness of the emulsion at each point of the negative was proportional to the intensity of the corresponding primary color that struck the film there. For example, if a particular point on the object had a lot of blue light, the emulsion of the film that records the blue intensity was thick at that point. Each film was then soaked in a dye that was the complement of its primary color, and the amount of dye that was absorbed was proportional to the thickness of the emulsion. The point which had a lot of blue, for example had a correspondingly thick emulsion in the film that recorded the blue primary color, and that emulsion absorbed lots of yellow dye. The three films could then be combined using the usual reversal process to produce a positive full-color image. This idea is the basis for the “Technicolor” process. This process is too complicated for everyday use, since it requires three separate cameras and a complicated processing scheme. Although later versions of the Technicolor scheme used only one camera, the system was still too cumbersome for most non-professional users.
Instead of using the thickness of the emulsion to control the contribution of one of the primary colors, an emulsion for a color film consists of silver halide with a chemical called a "coupler." The coupler is a chemical that reacts with the developer to form a colored dye wherever free silver is formed. There are two general techniques for introducing the couplers into the film. A substantive dye coupler is introduced into the emulsion layer during the manufacturing process. This idea is used in Ektachrome film. A non-substantive dye coupler is introduced during development. Kodachrome and Fujicolor are based on this technique. There are advantages to either technique.
Instead of a single emulsion of gelatin and silver halide, a color film has three separate emulsions -- one for each of the three primary colors. The top emulsion is a standard pure silver halide, which is sensitive only to blue light. In a substantive color film, this layer also contains a chemical that produces a yellow dye (the complementary color to blue) during the development process wherever there is also free silver. Thus the density of the yellow dye produced in this layer at any point is proportional to the intensity of blue light from the object that struck that point originally. In a non-substantive film, the yellow dye that winds up in this layer is introduced during the development process, as described below.
Since the following layers are also sensitive to blue light (in addition to the another primary color) the blue emulsion is followed by a yellow filter, which absorbs the blue but passes red and green. The next layer is usually green sensitive – as above, in a substantive film, the silver halide is combined with a coupler that produces a magenta colored dye, which is the complement of green. A non-substantive film has the magenta dye for this layer inserted during the development.
The final layer is the red sensitive layer, which contains the usual silver halide combined with an appropriate sensitizing chemical to make it red sensitive. This layer is combined with a cyan coupler (either during manufacture or development as above).
The developing process for a negative substantive film is similar to the process used for black and white film, but the chemicals are different. The developer acts on all of the layers simultaneously. It produces silver grains in each layer wherever light of the corresponding color was present in the original object. The silver combines with the coupler and the developer to produce a spot of the dye at that point. The color of the dye is the complement of the original color, so that the resulting negative is reversed both with respect to color and with respect to intensity. That is, a point where lots of red light struck the film will have lots of free silver in the bottom emulsion layer, which is red-sensitive. During the development process, the free silver combines with the coupler to produce cyan dye at that image point (the complement of red). (A cyan dye looks cyan in color because it passes cyan and absorbs red – just like the cyan inks used in four-color printing.) A point which had no red light from the object will have no free silver and no cyan dye. The unexposed silver halide and the silver produced during the development are washed away, leaving just the dyes formed in each layer. Points where no dye was formed are transparent in that layer. Thus if some point on the object produced only red light, the corresponding points on the blue and green layers of the film are clear, while the red layer has cyan dye at that point (which absorbs red).
A positive print is produced using a second exposure as for black and white prints. A white light is placed behind the negative, and the image is projected onto a second piece of film or photographic-sensitive paper. The point on the negative which has cyan dye as a result of the first development process passes cyan and blocks red. Since the other layers at that point are clear, the blue and green components of the light pass through the negative without attenuation. The light that passes through the negative at that point therefore contains all colors but red. When this light strikes the second photosensitive paper (which will be the final print), the silver halide in the blue and green layers are converted to silver, producing yellow and magenta dyes at that point; the red layer of the second exposure is not altered since the cyan dye in the negative removed the red light. The unexposed silver halide at that point will therefore be washed away at the end of the second development, leaving a clear spot in that layer. When the resulting image is viewed in white light, the yellow dye in the first layer absorbs blue, the magenta dye in the second layer absorbs the green, and red light is not absorbed by the clear third layer. The point looks red – just like the original point on the object. Conversely, a point that had no red light in the original object has no cyan dye in the negative and that point on the negative is clear. During the second exposure, the red light converts the silver halide in the red layer to silver, which combines with the coupler to produce cyan dye at that point. When that point is viewed in white light, the cyan dye absorbs the red component at that point, so that the reflected color has no red it. Thus a red point on the original object results in a point that appears red in the final image and a point that had no red in the original object has cyan dye at that point and does not appear red.
The same thing is going on simultaneously in the other layers. To summarize, after the first development, the negative consists of three layers of dyes, which are the complements of the original image both with respect to intensity and color. Bright parts of the original object have lots of dye in all three layers and appear black because of the usual 3-color subtraction rules. The same is true for all other intensities and colors. The second exposure and development of the printing process reverses the whole thing again and produces a positive image. Note that unlike printing, there is nothing that corresponds to the 4th black ink that printers use, so that we would expect that subtractive colors films would have difficulty reproducing an object that had lots of black in it. (Fortunately, such objects are not very common, but this fact explains why color film is not used to produce slides which have mostly black and white content, such as slides containing only text.)
The development process for a film based on the non-substantive process is similar, but the details are different. Since the film does not contain the dye couplers, these must be introduced during the development process. There are usually three phases to the development process – each of which is used to convert the latent image in one of the layers to free silver and introduce the appropriate dye coupler at the same time.
A color slide is produced using the analog of the reversal process described for black and white reversal positives. The first development converts the latent image into free silver and the dye of the complementary color. These products are then washed away and the film is exposed to light a second time. The areas that were not exposed to light from the initial object are now exposed, the remaining silver halide is converted to silver, and the silver interacts with the dye to produce the colored image. For example, a point on the object that produces only blue light converts the corresponding point on the blue layer of the film to silver and yellow dye during the first development. The remaining layers are not altered. The yellow dye and the silver are washed away leaving a clear substrate on all of the layers at that point. The film is then exposed again to white light. The red and green layers are now affected, producing cyan and magenta dye at that point, respectively. The blue layer is clear at that point, so nothing happens there. When the film is placed in a projector, the cyan and magenta dyes absorb the red and green light, respectively, and the blue portion of the light passes through the film and is seen on the screen. A point on the object that had no blue light does not affect the corresponding point on the film. The halide in the blue layer at this point remains unaltered until the second exposure converts it to silver and yellow dye. The silver is washed away leaving yellow dye at that point. When the film is projected, the yellow dye at that point absorbs blue, so that the projected image of that point contains no blue light, which is exactly the same as the original object. The other layers behave in the same way, so that the resulting image has the same color balance as the original object.
The sensitivities of the three layers in color films are adjusted to match the kind of light that will be used to illuminate the subject. For example, daylight is much bluer than incandescent light bulbs, so that a film designed to be used outdoors has its sensitivity to blue light reduced relative to the sensitivities at red and green. If this film is used to photograph a subject illuminated by normal incandescent lights, the resulting picture will look too yellow, because the reduced sensitivity to blue built into the film combines with the relatively small amount of blue light from the light bulbs to produce a picture that is lacking in blue and therefore looks yellow. Conversely, an “indoor” film designed for incandescent lights has its blue sensitivity enhanced to compensate for the relatively small amount of blue light emitted by incandescent light bulbs. It will produce photographs that are too blue if the subject is outdoors. Finally, although fluorescent lights appear to be white by eye, they contain lots of green and yellow (because of the very strong lines in the mercury spectrum at these wavelengths), and these strong colors will affect the color balance of all color photographs.
Black and white films also have a sensitivity that varies with the color of the incident light. Most black and white films are more sensitive to blue light than to light of any other color, so that objects that have lots of blue in them (like the sky) appear much brighter in a photograph than equally bright objects of other colors. Some black and white photographers use a filter in front of the lens to correct this imbalance. The filter is designed to absorb sky blue, and therefore has a yellow/orange color.
The dynamic range of a film specifies the range of light intensities that the film can handle. Although this value varies from film to film, most black and white films have a dynamic range of about a factor of 100 (that is, almost 8 f/stops). In other words, the brightest point on the image can be about 100 times brighter than the dimmest point. All points brighter than the brightest point cannot be distinguished – they all look white, and points dimmer than the dimmest point all look black. This dynamic range is a function of the film, but it can be moved around with respect to the brightness range of a complex object to overexpose or underexpose some part of the object. Color films have a somewhat smaller dynamic range because the exposure interacts both with the brightness of the image and with its color balance. The dynamic range is usually about a factor of 50 (that is, 5 or 6 f/stops).