Figure 2.1 shows a cross section of a color negative film. The first layer is sensitive to blue light. A yellow filter layer prevents any blue light reaching the green-light-sensitive and the red-light-sensitive layer, which are sensitive to blue light as well. When developing the exposed film, yellow, magenta, and cyan dyes are formed in the three layers. Photographic negative film produces its maximum amount of image dye from a maximum exposure and its minimum amount of dye from a minimum exposure. The result is an image where the tonal range and the colors are inverted.

The exposure of the negative depends on the luminance of the objects in the photographed scene. The absolute level of the luminance is compensated by choosing an appropriate combination of film speed and lens aperture. The relative luminance of the objects determines the tonal composition of the image.

The luminance of surfaces is measured in candela per square meter or foot lambert. Both are linear scales, that means a surface emitting or reflecting twice as much light as another has double the luminance. In the scene of Figure 1.1, the pathway (3) is approximately 16 times brighter than the deep shadows in front of the house (1). Some of the sun reflections on the roof (5) are again 32 times brighter than the pathway. The dynamic range, the ratio of the brightest to the darkest part in this scene, is approx 500:1.

Photographers and cinematographers prefer to measure exposure in “stops”. That is a logarithmic scale, in which a surface emitting or reflecting twice as much light as another is “one stop higher”. In other words, the pathway in Figure 1.1 is 4 stops higher than the shadows and some of the highlights are 5 stops above the pathway. The dynamic range of the scene is 9 stops.

Another scale for exposure, used in densitometry, is relative log exposure (log E). A change of one stop equals 0.3 log E. Table 2.1 gives a comparison of the different scales.

You do not need to learn those scales; important is to know the relation between stops, log exposure, and linear exposure.

Exposure value (EV) for a certain film speed and luminance (in fl or cd/m²) are absolute scales, while the other scales in Table 2.1 are relative. The reference point for the relative scales is arbitrarily set at EV 11. Any other value could be used as well.

If incident light with an intensity of one foot-candle (=10.76 lux) falls upon a perfect diffuse reflector then one foot-lambert (1 fl = 3.426 cd/m²) is reflected. If the reflectance of the object is 18% the luminance would be 0.18 foot-lambert. This is a rare example of an imperial unit system being easier to use than metric units.

A transparent material like film absorbs some of the light passing through it. The ratio of the transmitted to the incident light is called transmission or transmittance. For example, the illumination level is 200 units, and a piece of film placed between the light source and the sensor reads at 100 units: the transmittance is 0.5. Transmittance decreases with the darkness of the film.

Opacity is the reciprocal value of transmittance. It increases with the darkness of the sample, so it is a more convenient way of expressing blackness than transmittance, but its numerical value can range from 1 to 10,000.

Therefore, density (the common logarithm of opacity) is often preferred to transmittance or opacity. Like opacity, density increases as the film darkens. Density has the additional advantage that it accords with our perceptual judgments of how dark a sample is. In making a visual judgment, you would tend to describe a density of 1.0 as being about twice as dark as a density of 0.5. In this document density is written as “log D”.

With color negative film the density of each of the three dye layers has to be measured. This is done with a so-called “Status M” densitometer. Status M is an ISO (International Standards Organisation) standard specifying the spectral ranges of the visible light used to measure the three layers. Measuring “above base” means to subtract the density of the unexposed film. More information about Status M is given in the appendix.

Current color negative materials can capture a dynamic range of 9 stops (approx 500:1 in linear scale) or more. Have a look at Figure 1.1 again, where markers show five parts with specific tonal values.

Figure 2.2 illustrates how the negative reproduces the tonal range of the scene as film densities. Such a graph is called the characteristic curve of the film. Note the relation between the log exposure and the resulting density is linear over a wide range. The “toe” is the slow increase of density at the bottom of the curve. The maximum density of the highlights sits around 1.6 log D above base. If there were even brighter highlights in the scene we would see a roll off at the end of the density curve, which is called “shoulder”.

The slope of the straight part of the curve is known as the “gamma” of the film. The gamma of a typical motion picture color negative film is 0.6.

The numbers in the column titled “Label“ refer to the markers in Figure 1.1. Enlarged portions of the marked areas are displayed in Figure 3.7.

The cross section through an intermediate film looks similar to the structure of negative film (see Figure 2.1). Intermediate films are less sensitive than negative films and therefore have a much finer grain as one can see in Figure 2.3. The fine grain is desired because the duplication process should not introduce additional grain to the image. The density of intermediate films (both IP and IN) is measured with Status M.

Intermediate film is designed to have a characteristic curve that is linear over a range of 2.0 log exposure, so it can reproduce the density range of camera negative films; the gamma is 1.0. Since camera negative always shows grain, even in the unexposed areas (see Figure 2.4), the IN has to be heavier than the OCN. Otherwise, one would lose some of the grain structure in the shadows. As shown in Figure 2.5 the density in the IN is shifted by approximately 0.2 log D.

The printing process is monitored with the Laboratory Aim Density (LAD) method introduced by Kodak. The LAD is a patch with standardized density in the negative. This patch is printed to specified aim densities when making internegatives, interpositives or prints.

The laboratory controls print density and color balance by increasing or decreasing the intensity of each primary color of light in steps called printer points. Since the development of the Bell & Howell Model C printer most manufacturers have standardized on a range of 50 light points in 0.025 log E increments. Some laboratories, however, use other exposure increments.

In Figure 2.6 one sees the order of the dye forming layers in print film is different from negative film. The first one is green-sensitive, followed by the red-sensitive and blue-sensitive layer.

The density of print film is measured with a different spectral curve, named “Status A”, which is another ISO standard for densitometry.

The characteristic curve of print film is steeper and more S-shaped than the one of negative film; compare Figure 2.7 with Figure 2.2. The density range of print film is much higher, it is almost 4.0 for standard release print material. As a result the midscale gamma is around 3.

Like the IN/IP process the print process is controlled by the LAD method. The aim density for the LAD patch in the print is around 1.0.

Printing an internegative onto print film results in a relation between negative and positive densities that is illustrated in Figure 2.8. The normal placement of the LAD and of the 90% white, 18% gray, and 2% black tonal values is shown as well.

When a color negative or an IN is printed onto print film and projected, the resulting print-through curve, the relation between scene and screen luminance, is shown in Figure 2.9.

The gamma of the print-through curve has a midscale value of 1.8^[1]. Also, one can see a shadow and highlight compression in Figure 2.9. There are three reasons that the grayscale of the motion picture system is not one-to-one with the original scene.

Since the observer will adapt to the projected image itself, it is acceptable to reproduce a white at a density higher than the minimum density of the print film (look at the 90% level in Figure 2.9). This makes it possible to use some of the density range of the print film to reproduce information corresponding to luminance levels above white, like specular highlights. By comparison, reflection images, photographic paper prints for example, must reproduce whites at densities close to the minimum density because the typical viewing environment contains white objects that essentially control the brightness adaptation of the viewer. A reflection image with darker whites would look too dark.