Imagine, for a moment, if we could accomplish a statistical analysis on a digital photo.
We would desire some way to count how many dots of each color there are in a given picture. And, once we had that count, we would like to chart it somehow, so we could see at a glance how those dots compare against each other. This chart could reveal us whether our picture was washed out or too dark before we even printed it.
This chart is termed as a “Histogram”. Many digital cameras will create it for either the picture you’ve taken or the one you’re about to take. And, most photo editing programs will generate one, too.
When we watch the histogram of a picture, we are looking at the extent of color information contained in a photo. A dark photo will have much data on the far left side of the chart, while a “daylight” photo will be somewhere in the middle. The chart could reveal a single hump, or a series of spikes; it could be very tight, only a fraction of the histogram, or it could spread from edge to edge.
There is no such thing as a “perfect histogram” because every photo and every histogram is one of its kind, unlike anything else. But grasping how to read the lines on a histogram can render us an insight into whether or not we have the image we want.
There is actually only one kind of histogram that points out a bad picture, and that’s one where the data is up against the edge. The far outer area is pure black, and the far right outer area is pure white. If there is a big amount of absolute black or white in the picture, then some detail has most likely been lost–because it’s very rare to have pure, absolute black or white.