What is a Histogram? If you’ve got a digital camera, you’ve probably heard the term, maybe even you know what the histogram screen is on your camera, but do you know what it represents and how to use it?
A Histogram looks something like this:

This one looks like a nice silhouette of a mountain. The Histogram can be of all different shapes, and, as I’ll explain later, there is no ideal histogram shape.
Now you need to understand how to read this histogram. Basically the histogram is a vertical bar graph showing the lightness values of an image. On the left is pure black, on the right is pure white, and everything between is various shades of gray. Let’s assume you have an image that is 100×100 pixels square. Of those 10,000 pixels, we’ll assume 100 of them are completely black. So the vertical bar at the far left of the graph will represent 100 units on the graph.
Let’s assume the image has 20 pixels that are completely white. The vertical bar on the right of the graph will be 20 units high. All the other values of light work the same way. For those pixels that are just above complete black, their bar will be just to the right of the black bar. Those pixels that are nearly white, but not quite, will be just left of the bar representing white on the right of the graph. Pixels that have a middle ground in terms of brightness will have a bar somewhere near the middle of the graph.
Just to be clear, we are talking about brightness values, not color. Imagine the image was turned into a black and white image. It is those grayscale values we are measuring by in he histogram. Many cameras also have histograms for each color channel (Red, Green, and Blue) that graph values of each color in the image. So, with black being the absence of all color, each of the three RGB histograms would have pixels in the far left bar for each black pixel. White, being the full presence of all colors would also have a pixels in the right bar for each graph. But something like a nice purple color might have little or no pixels in the green graph, and varying amounts in the Red and Blue graphs depending whether the purple leaned more towards a blue-purple, or a magenta color.
I hope that clears up the concept of what a histogram physically is. Now, what do you do with it?
When you take a picture with your digital camera, you look at it on your nice little LCD display, and maybe it looks good there, but when you get home and put it on your computer, it might look a little dark or bright. Here’s the problem: LCD displays on the camera are not always calibrated well, plus you often have the ability to change the brightness of the LCD display, which might make an image look brighter or darker than it really is.
Enter the histogram…
When you preview your image in the LCD, don’t look at the image as much as looking at the histogram. It takes some practice to get used to reading a histogram, because, as I mentioned before, there is no one histogram that is right for every situation. If you are taking a close-up shot of someone’s face, for example, properly exposed skin tones will usually fall slightly high of middle on your histogram if the shot is taken in normal lighting. The histogram may look something like this:

But if you are taking a picture of a dark pine tree in a snowscape, the histogram is likely to have spikes in values near the low and high end for the darkness of the tree and the brightness of the snow, and short bars in the center for the lack of middle-ground values:

If you are shooting a picture of snow and your histogram shows a spike in the middle and nothing on the far right, there is a really good chance that your snow will look gray and dull instead of bright white and you will need to find a way to compensate and get more light to your camera’s sensor to make the image brighter or it will be a dark, underexposed picture. The same can be true of a picture shot in a dark nightclub. Your camera will try to make the dark nightclub average to a normal exposure, but to keep the mood of the image, you want it to look like a dark nightclub. If your histogram has a spike near the center and nothing to the left, you know your dark values are getting overexposed and bringing too much light into your shadow areas. You will lose the dark mood of the nightclub.
Be careful to watch for the histogram getting smashed up against either side also. If you have a large spike at the far right of your histogram, you have overexposed a large part of the image and washed it all the way out to white. There will be no detail left in those parts of the image. If you have a “Hightlights” screen on your camera where all the white areas flash black, this is showing the over-exposed parts of the image. This may be ok with you for the image you are shooting, but you need to understand it. Maybe you are shooting indoors and exposing the subject inside the house well, but the light coming through window is over-exposing what is outside. This could be intentional, and may be necessary to get the right exposure on your subject without using artificial lights.
If your histogram is smashed over to the left side, however, there is a lot of detail that has been lost to shadows. Using software to try to bring those details back will result in a very noisy image, and anything that is a full black pixel has no detail to recover. Again, this could be completely intentional. Back to the nightclub or other dark-lighting situation. It could be that your subject is a guitar player under bright lights on stage, but the background is dimly lit and fades off to black in your image. That might be the artistic look you are going for, and again, may be necessary to avoid over-exposing your subject without artificially lighting the background.
If your histogram is smashed up against both sides of the graph, then you are losing highlight and shadow detail somewhere. The contrast of the scene is more than the camera’s sensor can handle. This is a common situation when shooting in bright sunlight with the subject in shade, or just from the shadows cast by the subject themselves. For this situation, you need to decide what is most important… The highlights or the shadows. A bride’s dress next to a groom’s tux in bright sun always poses such a challenge. A little fill-flash might bring up the shadows enough to eliminate the underexposed portions of the image.
The histogram is a great, and crucial tool to use when evaluating your images on an LCD. In-camera meters cannot meter correctly for every situation. They are getting more and more advanced, but they still cannot guess about your artistic style or understand every lighting situation. The histogram is a great way to evaluate your image to make sure you are capturing the image you want without over- or under-exposing your image.
