Histograms – A Valuable Tool for Photographers
A photo histogram is basically a graphical representation of the light distribution in a photograph. In its simplest form, it's is an evaluation of light vs. dark areas within the drawing. There are more complex histograms that also analyze the distribution across a given color channel, but we won't get into those today. Histograms are commonly misunderstood or even ignored among amateur photographers. For that matter, histograms weren't available as a consumer tool until digital photography came onto the scene. Now, many cameras can generate an instant histogram from photos you're taking in the field. This has helped many photographers make quick adjustments on site or in the studio. In this article, I will give a brief introduction to Histograms, and I will explain how they can help you improve your own photography.
[singlepic=211,280,,,right]The histogram for a typical photograph will generally span the full spectrum with several peaks and valleys. The left side of the histogram represents the darker values of the image while the right side represents the brighter values within the image. The spikes are simply a measure of how many pixels within that range, the taller it is, the more pixels at that particular range. This histogram shown here at the right was taken from Adobe Photoshop. Other software will have something similar. Your camera's histogram may or may not show specific data – such as mean, median and the standard deviation – but its graphical representation will help you to better understand your photo's exposure.
Typically, a perfect exposure will result it something not unlike this example. It will have at least one major spike somewhere near the center of the histogram. Subsequent spikes are a result of other dominant colors and or value ranges. Additionally, a perfect exposure will not have any clipping (overflow) at the edges. Clipping is commonly indicative of extreme overexposures and underexposures. As you already know, overexposures and underexposures will cause a loss of detail in those portions of the image.
Before I go on, I need to clarify one common misconception. Many photographers use the term, perfect exposure, to indicate that a photo has an even value distribution. Don't let this mislead you. A perfect exposure is not always the goal. There may be nothing wrong with an image that has a flat profile. There may not even be anything wrong with a histogram that is clipped on either end, especially i you're working with high-contrast black and white photos. Remember, the histogram is only a tool to help indicate potential problems with your photo. As the artist, it is up to you to correct or ignore any flaws that may be indicated by the histogram.
[singlepic=209,320,,,left]Low-Key and High-Key Histograms
An image like this one is called “low-key”. Most of the image's values fall in the lower end of the histogram with very little in the upper range. In this particular case, when I took the initial shot, the camera made some automatic adjustments so as not to overexpose the sun itself. The result was an image that was incredibly dark. I was expecting a low-key image, but I wanted some detail of the leaves and I wasn't so concerned if I lost detail in the sun (it is a light source after all). After viewing the image histogram, I retook the shot with a few adjustments so that the exposure time was a little bit longer. In this final version of the image, the sun itself is slightly overexposed, which is why the histogram clip is indicated on the right side of the histogram. But as I said, I wasn't concerned about that. Instead, my exposure adjustments allowed me to avoid losing too much detail at the dark end of the histogram. Without these adjustments, I wasn't able to clearly distinguish between the water and the island. The histogram helped me figure out why and how much of a tweak I needed to make to my settings. (View the full-size image: Sunset on Pocono Lake).
[singlepic=210,320,,,right]The photo example at the right is referred to as a “high-key” image. As you would expect, the high-key histogram is heavily weighted at the lighter side (right side) of the histogram. High-key histograms. With this shot, I purposely wanted to focus on the lines created by the roof, and I purposely overexposed this image. While this isn't the norm, there are certain cases where an overexposed image may be appealing, and a histogram can help you get there.
A histogram also gives you a pretty good indication about the contrast in a given photo. If all of the peaks in a histogram are grouped together in a narrow area and the left and/or right side of the histogram is flat against the bottom, there is not much contrast in that image. A high-contrast image will have a histogram that pretty much fills the entire range from left to right. The peaks of the histogram will be more rounded at the top. Again, this is only for informational purposes. Depending on your goals, you may not want a lot of contrast, or you may want a lot.
In any case, the histogram can be used to help the artist achieve his goal. Getting the exposure to meet your needs is what photography is all about. When setting up for a shot, you will have some concept of how you want the final shot to appear. Tweaking the shutter speed, ISO or even adjusting the light-source will help you to achieve your goals. And while the histogram won't give you all the answers, it will at least help you make adjustments efficiently. But the histogram is completely useless unless you have a goal for your shot. So the first step in using a histogram is to make sure you clearly define your goals for a given shot.
The item that I cannot stress enough is that there is no such thing as an ideal histogram. The histogram is only an indicator of your photo's value range. Your goal is not to get an ideal histogram. Your goal should be to achieve a histogram that matches your intentions. After working with them for a while, you should be able to predict what the histogram will look like based on your camera's settings. And that is when you know that you're using the histogram most effectively as a tool.