It takes a little practice, but you’ll soon be dialing in accurately balanced color for every image in just a few minutes. Here’s how:
1. Click and hold the Eyedropper in the Tool palette and select Color Sampler Tool from the flyout menu.
2. Set sample size for the tool to something larger than Point Sample, since any given pixel could be off. The larger the resolution of the file, the larger the average you can use. This is a 5-megapixel file, and I’m using the 5 by 5 Average setting.
3. Click somewhere on the image that should be a neutral gray.
4. Check out the Info Pane and the RGB values there. These are the values of the point we just sampled. For a perfectly neutral gray, all three values will be identical. Notice that while Green and Blue values are relatively close, the Red channel is much lower, which makes sense considering the overall cyan cast of the image.
5. To fix this, we need to bring the Red value up. Start by adding a Curves Adjustment Layer (Layer > New Adjustment Layer > Curves, or by clicking the Curves icon in the Adjustments Pane).
6. Next, we’ll be using the Properties Pane. Switch to that pane and click the onscreen adjustment option (the icon that looks like a finger pointing between two arrows). This allows you to click and drag up or down anywhere on the image to increase or decrease the value of the selected channel. I need to bring the Red value up, so I’ve selected the Red channel from the RGB dropdown.
7. Now click and hold anywhere on the image while you drag the up or down to adjust the value of the Red channel. My point sample had values of R=43 G=73 B=67, so I dragged upward until the Red channel had the same value as the Blue channel, R=67.
8. To get all channels to the same value, I next selected the Green channel from the RGB dropdown and dragged downward, to bring the Green value down to 67. With all channels now at the same value, you’ll see that the image has a much more neutral color balance.
9. To get the balance even more precise, try taking a sample from another area of the image that should be neutral. My first sample area was in a midtone area, so I placed two additional samples in shadow and highlight areas and balanced them using the same method. Note that when adjusting additional points, your original point may be affected. That’s fine as long as the RGB values stay within a range of something near +/- 5. You can also go back and readjust the previous points and usually get them closer, if not precisely equal, as I did here.
Compare the final image with the original and you can see how powerful this relatively simple technique is for achieving accurate color balance. What I like most about the approach is that you’re not relying on guesswork or judging by “look”, nor will your eyes be fooled if your monitor’s calibration is off, because you’re making your changes based on the numerical values of each channel. It’s a foolproof way to arrive at perfectly neutral color.