Even a colorblind person can color-correct a photo
-or-
How to correct your color using RGB values
by Chris Nicholson

Part I: Prep Work

Several times people have asked me, "How can you be a photographer if you're color blind?"

Cape Blanco Light, Oregon. Nikon F5, Nikkor 35-70mm f/2.8. More Lighthouse Pictures.

This question comes after these people learn that my eyes are particularly insensitive to the red/green areas of the spectrum. Sure, I can tell the difference between a red flower and green lily pads; I see red and green, just not the same way most of the human race does.

Still, color perception has nothing to do with how you capture an image through a lens; you see what you see, I see what I see, and if we know how to make a good exposure, then everyone's happy.

The only problem with a colorblind photographer comes in making a print. When I was in college I knew I could never be a printer in a lab because I wouldn't be able to accurately color-correct an image.

Alas, in the digital age, that's no longer true. Anyone who knows how to read numbers can color-correct an image with a computer. Here's how.

As an example, we'll start with this unadjusted scan of Oregon's Cape Blanco Light. Obviously, it needs to be color-corrected.

It's a black, white and gray issue

Almost every image will have a black point and a white point — that is, a place in the photo that should be absolutely black (usually deep in a shadow) and another that should be pure white (a bright cloud, water rapids, etc.). You need to find those two points and determine the RGB value of them.

First, do a quick Levels adjustment. Move the sliders in to just barely clip the ends of the black and white ends of the histogram. Some people like to clip off more information in Levels, but I don't like to blindly throw info away; if I want to throw out more info later, then I can do it then.

In the Levels dialog, move in the sliders for the black and white ends of the histogram (the far-left and far-right sliders). You want them to just clip the ends of the histogram graph.

Next you need to find where in the image your black and white points are. An easy way to do this in Photoshop is to navigate through the menus to Image-Adjust-Threshold. Move the slider to the far left to see the black point, to the far right to determine the white. (You may need to nudge the slider back in from the edge to see the points.)

Cancel out of the Threshold dialog and position the mouse pointer over your black and white points. Read the RGB value for each point in the Info box, and write the values down. (Alternatively, with the Eyedropper Tool, you can shift-click on each point, which will lock the values into the Info box.)

With the Eyedropper Tool we've marked the black (2) and white (1) points for future reference.

Next, try to find something in the photo that should be gray - snow in shadow, faded cement, a gray shirt, etc. Now record the RGB value for this gray point.

There's nothing in this photo that strikes me as needing to be gray. However, I know that this point (3) on the lighthouse is white, but it's not being hit directly by the sun, and therefore won't be pure white. Should it be pure gray though? Perhaps not; we'll find out later.

(The gray point can be a lot harder to find than the black and white points - in fact, in some cases, there simply may not be any pure-gray elements of an image. In these cases, just use the black and white points.)

Rather than writing down the RGB values, I shift-clicked the Eyedropper Tool on my black, white and gray points, which locked in their RGB values in photoshop's Info box. This makes it easy to refer to them in the following steps.

Part II: Correct the Color