Bruce Barnbaum - The Art of Photography - An Approach to Personal Expression

The Sensor’s Useful Brightness Range

Each sensor has its inherent brightness response range (referred to as the dynamic range). The range varies from one camera model to the next and may very with ISO setting. Just as with film, the digital sensor requires a minimum level of light to register shadow values (equivalent to the exposure threshold for film). All brightness levels in the photographic subject that fall below the sensor’s response threshold will be depicted as black. (This is similar to the film negative having areas exposed below threshold, having no density, and therefore supplying no information for the print. For transparency film, it’s simply getting unexposed black areas). In addition, at low brightness levels (equivalent to exposures slightly above the film threshold), random electrical signals generated by the sensor and related circuitry will constitute a significant portion of the information and can be expected to appear as noise. At the highlight end of the sensor’s range, above a maximum brightness level, the sensor will “clip” the highlights, that is, it will not differentiate additional brightness (equivalent to a pronounced shoulder on film) and all brightness beyond the dynamic range will appear as blank white. Clipping is analogous to overexposing transparency film to the point where the highlights are rendered as clear film base.

In the early days of DSLR cameras, the dynamic range was similar to that of most outdoor color transparency film, a range of approximately 5 f-stops. Current DSLRs have a dynamic range of as many as 10 f-stops or more using RAW capture, the camera’s base ISO, and no image adjustment, thus equaling, or perhaps even exceeding Fujichrome 64T transparency film, but falling far short of the useful range of color or black-and-white negative film. The useful dynamic range may vary with the ISO setting and can be expanded to some extent in both the shadows and highlights using the shadow and highlight adjustments in the RAW converter. Determining the dynamic range of your camera is discussed in the next section.

The brightness information from each photosite is converted by the camera’s circuitry from analog output to digital data, which quantifies brightness in a geometric progression. That is, the output from a photosite that receives an exposure of up to one f-stop above the threshold will contain one bit of binary data representing one of two possible responses—black or the first brightness level of red, green, or blue (depending on the color filter over the photosite) that is lighter than pure black. The output for the same photosite that receives up to 2 f-stops above the threshold will contain 2 bits of data representing any one of the next 4 lighter shades of its color. The output for the same photosite that receives up to 3 f-stops of exposure above the threshold will contain 3 bits of data representing any of the next 8 lighter shades of its color. The progression continues so that a photosite that receives an exposure within the last f-stop of the dynamic range will record any one of several thousands of shades of its color from very light to white.

The photographic importance of the geometric nature of digital output is that exposures made at the first stop or two above the threshold will contain larger tonal jumps than will an exposure of the same subject made with more exposure. With each f-stop increase in exposure, the number of tonalities doubles and soon there comes a point where neither the printer nor the human eye can distinguish between the increased number of smaller tonal steps. Additionally, as brightness increases, electronic noise becomes increasingly less visible. Thus, up to the point of clipping, more exposure results in smoother tonal transitions, better color fidelity. More exposure also provides a reserve of data with which to make major tonal modification in your RAW converter or image editing software without posterization, and produces cleaner shadows. These are all desirable results obtained at the minimal expense of maximizing your exposure.

Technical explanations aside, the practical point is this: Give as much exposure as you can without clipping the highlights. The exposed capture will result in more data and better print quality. Lastly, don’t worry if your capture looks washed out on the camera display; you can easily preserve shadow detail and dial back the effect of the increased exposure using the settings in the RAW converter (to be discussed shortly).

While this isn’t exactly analogous to my recommendation (maybe my demand) to expose the shadows in Zone 4 in traditional photography, it’s close. In traditional (i.e., classical or film) photography, I’ve stressed the importance of exposing the shadows in Zone 4 to achieve better separations, even if you want them printed in Zone 3. Here, you want to push the exposure higher to get smoother, better, more detailed information. More information—particularly more shadow information, where it decreases with a lower exposure—means tonally richer images, with better gradations from one dark tone to the next. The limit to exposure is the dynamic range of the sensor. As long as you do not increase exposure to the point where the highlights are clipped, the result will be shadows with more tonalities, smoother tonal transitions, and less noise. The histogram discussed in the next section is your onboard tool for getting the exposure just right.

The importance of smooth tonal gradations cannot be overemphasized. Years ago I was shown a comparison of a classical (darkroom) print with a digital print derived from a scan of the original negative. From a distance they looked reasonably similar. But close inspection revealed disturbing artifacts in the digital print. The photograph contained dark bushes along a winding road in fog. In the straight classical print those bushes had minor tonal gradations that made sense. The digital print, however, had random blacks and grays intermixed that made no sense. It would be difficult to describe the difference, but perhaps this will suffice: If you’ve ever painted a wall in a room with a roller, you know how it gets spotty as the roller gets dry (that’s when you dip the roller back in the paint trough for more paint). Now try to imagine vertical strokes with a relatively dry black paint roller and horizontal strokes with a relatively dry gray roller directly atop it. That’s how the bushes looked in the digital print. But that was years ago. Today, working in higher bit depth and with proper exposures, you can avoid that type of noise.

The Histogram—The Heart of the Digital Zone System

The histogram is a graphic representation of the distribution of brightness within a capture. The histogram can be displayed on the camera, in the RAW converter, and in the image-editing program. The lowest brightness level appears on the left edge and the brightest level appears on the right edge of the display. A typical histogram for a properly exposed capture appears in Figure 11-3.

Figure 11-3. Histogram of a good exposure

For any capture, if the endpoints of the histogram do not extend to the left and right boundaries, the brightness range of the scene is less than the camera’s dynamic range. If the left edge of the histogram touches the left boundary, some of the lowest brightness levels in the scene fall below the sensor’s threshold and will be recorded as empty black. If the right edge of the histogram touches the right boundary, some of the brightest portions of the scene exceed the sensor’s dynamic range and will be recorded as empty white. If you reduce the exposure (i.e., use a higher shutter speed, smaller aperture, and/or lower ISO), the histogram will shift to the left; if you increase the exposure, the histogram will shift to the right.