Deciphering Histograms

No single technique or tool will improve your photos on its own, but mastering the use of many techniques and tools can certainly help you together. Learning to use and correctly interpret a histogram is just one piece of the puzzle when it comes to digital photography. It can help you figure out how to expose for difficult situations, and reduce the amount of images you end up tossing in digital post-processing. For those of you who shoot film, don't feel left out; you can still take advantage of the histogram on a digital camera if you use it to check your exposure before taking a shot. Whatever the medium, a digital histogram can be a useful tool when it comes to getting correct exposures.

How To Read a Histogram

A histogram is simply a visual tool used to shows you how the light is distributed in your image. It measures the different types of light, from shadows to highlights. You can split the histogram roughly into thirds to indicate a general range of the three types of light: the left side of the histogram is for darks, the middle for midtones or grays, and the right side is the highlights. If your image doesn't have many dark areas in it, then you would expect the graph to lean more to the middle and right of the histogram.

You typically would like the far left and right edges of the histogram to trail down to the bottom, and not just cut off. When the curve seems to go off the edge of the graph, the data gets "clipped" which means that you have either too dark darks or too bright highlights. Of course, depending on the effect you are aiming for, you may want one or the other (or both) edges of your histogram to be clipped. Typically, you would prefer the edges to taper down instead of being cut off. When the right side of the curve is clipped, your image is typically overexposed. When the left side is clipped, the image may be underexposed. 

Maybe it's best if you can see an example of a black and white image so you can clearly see what it looks like. Here is a black and white image and the corresponding histogram as interpreted by Adobe Photoshop:

You can clearly see that the image doesn't have much in the way of black, and is mostly midtone grays with some highlights in the clouds. Therefore, the histogram tells you the same thing, that you will get a very even image with plenty of midtones. What is nice about this histogram is that you can tell that the highlighted area has not been blown out too much. There are some peaks in the highlight section of the histogram, but the edges do taper off before reaching the edge of the graph, a good sign that data has not been clipped.

The most typical image that you might take would indicate that you would want some darks, some highlights, with the large bulk of the light in the midtones. However, that is only the most typical of images. Depending on what you are trying to capture, you may want something completely different in your histogram. What matters more is not getting an "ideal" histogram, but deciding whether or not the histogram that you are seeing is what you want to get out of your image.

A Few Examples

The easiest way to get the hang of reading a histogram is to know what the histogram looks like for a bunch of images. Remember, the key things to look for are: a graph that tapers on the left and right side, and peaks in the areas of the graph where you would like them to be (on the left side for more darks, on the right side for more highlights, in the middle for more midtones). Here are a few examples of both black & white and color images along with their histograms to help you get an idea of what to look for.

 You can see in this example that are a good amount of darker midtones and highlights, but not many right in the middle. You should also note that the edges of the histogram taper down and are not cut off at the edges.

You can see in the histogram that this image is heavy on the darks, and you can clearly see that the entire background of the photo is mostly black. When you are shooting a high contrast situation like this, you may want the image to be sharp with lots of darks and highlights and not much in the middle, or you may want it to look very dark or very bright. This image may be classified as being underexposed, since the background contains a good amount of very dark shadows that lack detail; you can see this in the very left edge of the histogram that is slightly clipped.

According to the histogram of this image, it's slightly over-exposed. You can see that the inner white curve just on the inside of the cup has been blown out to white, which is probably what the histogram is reading in the highlights area of the curve. It looks like there is plenty of room to shift the histogram to the left, so if it had been shot with one stop less exposure, the highlights would not have been clipped.

In a Nutshell

So let's quickly go over what we covered about how to read a histogram. The histogram is a visual way to interpret the exposure for an image. It can be split into thirds: the left side is the darks, the middle is the midtones, and the right is the highlights. The edges of the curve should taper down like a hill to the bottom of the graph on either side, or you have underexposed or overexposed the image and clipped data.

Links for more reading

PlanetNikon's tips for reading and understanding your histogram

Steve Hoffman's Practical Guide To Interpreting RBG Histograms

An article on histograms from Basic Digital Photography

How to Use Histograms by Ken Rockwell