Understanding histograms

Comparing photos and their histograms - all photographs were taken by myself.

Five photos and their histograms of Derek Falls in Manning Park, BC.

Understanding histograms is an important tool in understanding digital sensors and files.  Histograms are also important when editing files, as it gives an empirical reference to your work.  This is especially true if your monitor is not properly calibrated.

I use histograms in the taking and editing of my photos.  I often use raw files, both on my cell phone cameras and DSLR and mirrorless cameras.  The histograms allow me to interpret exposure, extract details from shadow and highlight areas that would otherwise not be present, and to play with contrast and other variables offered in many pixel-editing programs.

The top image - four different photos - shows different images where the exposure is correct (with minimal adjustments in Photoshop) in variable environments.  Each photo will have its own histogram, and you will have to interpret it, keeping the nature of the photo in mind.  A lot of dark values (concert) will produce peaks on the left side of the histogram, possibly against the left boundary as it is too dark to record details.  A lot of light values will produce a horizontally rotated histogram in the same way.  If I am shooting into the sun, it is a likely result.

Histograms can be bell-shaped; I would expect this in an evenly lit situation with a lot of mid-tones and fewer dark and light values towards either end of the curve.  Light and dark areas will increase the height of the curve in both respective areas.  To keep from over- or under-exposing an image, you have to pay attention to how the graph looks.  You can see examples of what I mean in the second image above.

As an image becomes more and more underexposed, the curve presses up against the left wall with the right side becoming increasingly vacant.  The opposite is true for underexposure.  The trick is in paying attention to all three areas of the curve; the left, the right, and the middle.  In extreme situations, you are going to have pixels blown out on both ends.  At that point, you definitely want to use a raw file type, balance the middle tones according to what you see, and hope for a lot of latitude.  Bracketing gives even more options, especially if you plan on using HDR software later.

Thanks for reading.

Eric Svendsen     www.ericspix.com