Color channels are often thought to be the exclusive realm of mystics and Photoshop gurus. If you are willing to dedicate a few minutes of time to learning, I’ll take the mystery out of channels, and give you the power to improve your workflow and your end results.
The colors we see on our monitors and in print are created by combining specific amounts of either Red, Green, and Blue, or Cyan, Magenta, Yellow, and Black with the result being a new intermediate color. Since the majority of our readers are using the RGB model, I’ll stick with that for our examples. If I get requests in the comments below, I’ll add a section explaining CMYK.
Most users of image editing applications like Photoshop or Gimp, as well as users of other graphic design applications are familiar with, or have heard mention of the 256 levels used to define color and density. Most often these levels are represented by numeric values from 0 to 255 with zero being one of the levels. In the RGB model, these levels can be equated with visual light, zero being no light, or pure black and 255 being maximum light – pure white.
When we build colors in the 8-bit RGB model, we are using 256 levels of Red, Green, and Blue in various combinations called a “build”. You can think of the color-build as a recipe for that specific color.
Collectively the color channels are nothing more than a representation of those recipes. And when the recipes for all the pixels are put together in the right order, we have our color image. Viewing our color channels is effectively changing the way your cook-book is organized. So rather than finding the recipe for the pixels on one page of your cook-book, your color cook-book has three pages, one each for Red, Green, and Blue. The Red page tells you how much red to use and where, the same goes for the Green page and the Blue page. So in our example above if we assume that each colored square represents 1 pixel, the Red page would tell us the first pixel would have 255 red, the second pixel would have 68 red and the third pixel has 126 red. The Green page would read: 1=128 and 2=68 and 3=0 and so on for the Blue page.
Photoshop shows us these channels in a way that our minds can easily process: as images. We can grasp the concept of images much easier than looking at the potentially millions to billions of numbers required for single image. Photoshop’s default is to show you these images as various shades of gray (256 possible shades to be exact). Here is what our example above looks like as color channels:
Where the build calls for zero of a color, that channel represents the area as black. Where it calls for all of that color, it is represented in the channel view as white. All intermediate values show up as the appropriate shade of gray.
Real World Examples
This image is pretty much straight out of a raw conversion. The file has been optimized in the conversion to make sure that none of the channels contain either pure black or pure white. This is to mimic the way the eye naturally sees. We’ll compensate for its somewhat flat appearance when we show you how to optimize your files without damaging your color fidelity.
Here is the view of the red channel, remember lighter areas indicate more red, darker indicates less:
Here are the Green and Blue channels, you can click them for larger viewing:
Notice that the lighter areas of the scene show as lighter in all three channels, and the darker areas of the image are darker in all three channels. You can also see that the areas of the image that are green show as brighter in the green channel in relation to the other two.
Also, all three channels have complete detail from shadows to highlights, nothing is lost. This is critical for full color fidelity. This full range of detail is essentially how channels should be. When channels look muddy or if there is “clipping” to full black or full white, there is a loss of color fidelity. I use channel views regularly to examine the state of a file’s “health”. If the file’s channels are not right, then I know right away I can’t generate the best possible print.
It is key to understand that in an RGB color space, a channel is both color and density information. Any change that you do to a channel will affect the color, saturation and density of your file. If you increase any value in a color channel, let’s say moving the red value of an area from 180 to 185, the resulting color will be more red and lighter.
See, no mystics required.
Reach out in the comments below with questions and comments.