THE VALUE OF THE HISTOGRAM
Simply put, histograms are two-dimensional representations of your images in graph form.
There are two different histograms that you should be concerned with: the luminance and the color histograms. Luminance is referred to in your manual as “brightness” and is most valuable when evaluating your exposures. In Figure 1.1, you see what looks like a mountain range.
The graph represents the entire tonal range that your camera can capture, from the whitest whites to the blackest blacks. The left side represents black, all the way to the right side, which represents white. The heights of the peaks represent the number of pixels that contain those luminance levels (a tall peak in the middle means your image contains a large amount of medium-bright pixels).
Looking at this figure, it is hard to determine where all of the ranges of light and dark areas are and how much of each I have. If I look at the histogram, I can see that the largest peak of the graph is in the middle and trails off as it reaches the edges. In most cases, you would look for this type of histogram, indicating that you captured the entire range of tones, from dark to light, in your image. Knowing that is fine, but here is where the information really gets useful.
When you evaluate the histogram that has a spike or peak riding up the far left or right side of the graph, it means that you are clipping detail from your image. In essence, you are trying to record values that are either too dark or too light for your sensor to accurately record. This is usually an indication of over- or underexposure. It also means that you need to correct your exposure so that the important details will not record as solid black or white pixels (which is what happens when clipping occurs). There are times, however, when some clipping is acceptable.
If you are photographing a scene where the sun will be in the frame, you can expect to get some clipping because the sun is just too bright to hold any detail. Likewise, if you are shooting something that has true blacks in it—think coal in a mineshaft at midnight—there are most certainly going to be some true blacks with no detail in your shot.
The main goal is to ensure that you aren’t clipping any “important” visual information,and that is achieved by keeping an eye on your histogram. Take a look at Figure 1.2.
The histogram displayed on the image shows a heavy skew toward the left with almost no part of the mountain touching the right side. This is a good example of what an underexposed image histogram looks like. Now look at Figure 1.3 and compare the histogram for the image that was correctly exposed. Notice that even though there are two distinct peaks on the graph, there is an even distribution across the entire histogram.
Simply put, histograms are two-dimensional representations of your images in graph form.
There are two different histograms that you should be concerned with: the luminance and the color histograms. Luminance is referred to in your manual as “brightness” and is most valuable when evaluating your exposures. In Figure 1.1, you see what looks like a mountain range.
 |
| FIGURE 1.1 This is a typical histogram, where the dark to light tones run from left to right. The black-towhite gradient above the graph demonstrates where the tones lie on the graph and would not appear above your camera histogram display. |
The graph represents the entire tonal range that your camera can capture, from the whitest whites to the blackest blacks. The left side represents black, all the way to the right side, which represents white. The heights of the peaks represent the number of pixels that contain those luminance levels (a tall peak in the middle means your image contains a large amount of medium-bright pixels).
Looking at this figure, it is hard to determine where all of the ranges of light and dark areas are and how much of each I have. If I look at the histogram, I can see that the largest peak of the graph is in the middle and trails off as it reaches the edges. In most cases, you would look for this type of histogram, indicating that you captured the entire range of tones, from dark to light, in your image. Knowing that is fine, but here is where the information really gets useful.
When you evaluate the histogram that has a spike or peak riding up the far left or right side of the graph, it means that you are clipping detail from your image. In essence, you are trying to record values that are either too dark or too light for your sensor to accurately record. This is usually an indication of over- or underexposure. It also means that you need to correct your exposure so that the important details will not record as solid black or white pixels (which is what happens when clipping occurs). There are times, however, when some clipping is acceptable.
If you are photographing a scene where the sun will be in the frame, you can expect to get some clipping because the sun is just too bright to hold any detail. Likewise, if you are shooting something that has true blacks in it—think coal in a mineshaft at midnight—there are most certainly going to be some true blacks with no detail in your shot.
 | ||
| FIGURE 1.2 This image is about two stops underexposed. Notice the histogram is skewed to the left. |
The main goal is to ensure that you aren’t clipping any “important” visual information,and that is achieved by keeping an eye on your histogram. Take a look at Figure 1.2.
 |
| FIGURE 1.3 This histogram reflects a correctly exposed image. |
The histogram displayed on the image shows a heavy skew toward the left with almost no part of the mountain touching the right side. This is a good example of what an underexposed image histogram looks like. Now look at Figure 1.3 and compare the histogram for the image that was correctly exposed. Notice that even though there are two distinct peaks on the graph, there is an even distribution across the entire histogram.
