web analytics

cg » a big easy photo blog

Masthead header

Lesson 2.1: Histogram

The instant feedback of digital photography has made the process of learning to take great pictures infinitely easier. Did I get everybody in the frame?  Check the LCD.  Did anybody blink?  Check the LCD.

Is my picture too bright/dark?  This is where checking the back of the camera can get beginners into trouble.

It’s important to understand how to read a histogram. Looking at the LCD screen on the back of the camera might not render an accurate picture (your camera’s LCD screen might be set to maximum brightness, for example, which can give you the impression that your pictures are washed out or overexposed) and the LCD screen is also a very small representation of the big picture (pardon all puns.)

A histogram is a graph that displays brightness values. It’s divided into five equal sections moving from left to right: very dark | dark | 30 percent gray | light | very light. The center of the histogram, 30 percent gray, is the value most camera manufacturers use as the default tone for metering and calculating exposure.

Let’s say the building blocks of a histogram begin with a rectangle drawn on a piece of paper. Only instead of looking at it straight-on, you’re going to view it eye-level to the table so that the graph becomes a vertical axis (and shows nothing at this point.) Then you’re going to use a box of tiny little square tiles from Home Depot. Each individual tile represents one value of brightness in your image. You start placing tiles representing pixels into the graph, stacking them according to lightness or darkness. As you get more than one of the same value, you begin stacking them.  You will find that images have a LOT of one value, less of others, so some of those little tile stacks will rise pretty high, maybe even to the top of your graph, others will be short. After all the tiles are stacked from the image, you’re looking at a histogram (vertical axis graphic representation of light and dark values.)

Generally speaking, the goal in a properly exposed image is to find most of your tiles stacked heavily in the middle and tapering off into very dark (left) or very light (right) edges. (Of course, this rule was made to be broken; first learn the rules, then learn to break them.)

Take for example this image of David and Shelley’s thank-you cards.  I’ll consider this image “properly exposed” because it’s a straightforward exposure using the exact settings my camera’s light meter told me to use (I had it set to use center-weighted metering; more on that later.):

Nikon D700, 135mm lens, f4 @ 1/250 second, ISO 200

Nikon D700, 135mm lens, f4 @ 1/250 second, ISO 200

Here’s how all of the individual tiles stack up to represent this image in my histogram:

Notice that there’s a good representation of tiles stacked across the histogram, heaviest in the center and stretching out to either side? This particular image is slightly heavier toward the right as there are a lot of lighter tones in it. (I also blew out the front of the cream colored table the cards were resting on. I did this intentionally, I was metering for the cards, my main subject, not so much the table or wall behind them.)

Here’s an example of this same scene, image intentionally underexposed:

Nikon D700, 135mm lens, f4 @ 1/3200 second, ISO 200

Nikon D700, 135mm lens, f4 @ 1/3200 second, ISO 200

Look at the histogram now and how all my little tiles are stacked heavily to the left; there’s no mistaking that no matter how the image appears on my LCD, it’s in fact very dark; also look at the cream table, which is closer to properly exposed then it was above (the !!! warning sign is still there, just disregard it for now. If you are working in Photoshop and come across it, it’s warning you that the histogram you’re seeing is cached from memory and you’ve since made adjustments to the image that might change its appearance. Just click it and Photoshop will generate a new histogram and the warning will go away.):

histogram of underexposed image

And the same scene intentionally overexposed:

Nikon D700, 135mm lens, f4 @ 1/30 second, ISO 200

Nikon D700, 135mm lens, f4 @ 1/30 second, ISO 200

And its resulting histogram:

overexposed histogram

overexposed histogram

Something interesting that I didn’t intend to find when shooting this example: I avoided most of these exposure issues in post processing by shooting in RAW. As I imported the underexposed and overexposed images into Lightroom, they appeared pretty similar to the correctly exposed image. I had to go back and manually un-do all of my regular workflow develop settings to get them to look like they did on my LCD as I was shooting.:)

[return to the PHOTO 101 Table of Contents]

we hate spam, too! your email is never published, sold or shared. required fields are marked *

*

*