Channels

page one
What are channels, and why should you care? Because they give you the power to control the degree to which pixels are affected by everything you do.
     When you open any color image in Photoshop, each color in the image will be represented in the Channels palette by a color channel. For example, RGB images open with a Red, Blue, and Green channel, as well as a composite RGB channel. We’ll discuss them later.
      First let’s consider the channels you create by saving a selection, or clicking the New Channel button in the Channels palette. These are called alpha channels. Alpha channels are saved masks. Masks are saved or stored selections. Selections are maps that tell Photoshop how much to apply color, effects, filters, or adjustments that you make while that selection is active. Understand that masks and selections are different forms of the same thing . Selections are temporary masks. Masks are stored selection maps. Alpha channels don’t show up in your image. They are stored selection maps which you can edit, and reuse as many times as you like. They are saved with the image so long as it is in Photoshop’s psd format.
     An alpha channel is a black and white ‘map’ of your color image. Where it is black, when loaded as a selection, it tells Photoshop, “leave these pixels unaffected by any changes.” Where it is white it tells Photoshop, “make these pixels 100 % affected by any color, filters, or adjustments applied.” Where it is gray, it tells Photoshop to affect those pixels in the same percentage as its shade. Lighter gray gets more affected; darker gray gets less affected.
     To show you that selections and masks are the same thing, try the following demonstration. Choose File > New to make a new white document. Create a new layer by clicking the New Layer button in the Layers palette. Make sure this new, empty layer is selected.
     Select the Type tool. In its options bar, choose the Masked Type tool (the T with a dotted outline, second from the left), set its options to Helvetica, Bold, 72 pt, and Strong (for anti-aliasing).

Type a capital A. With the letter’s selection outline active, choose Edit > Fill. Click the color square in the Fill dialog box to pick a color. I chose a bright red for my color; you can use black if you prefer.
     Next, go to the Channels palette. Click the New Channel button at the bottom of the palette. Your white document will turn completely black. This is simply the new channel which is a mask covering the entire image. Choose the Mask Type tool again with the same settings. Click on the black mask, and type a capital A. Choose Edit > Fill and pick White from the menu. Choose Select > Deselect, or press Ctrl-D to deselect the white A on the mask. In the Channels palette, with the new channel still selected (showing dark blue) click on the Load Selection button, or go to Select > Load Selection and choose Alpha 1 from the dialog box’s menu. Click on the composite RGB channel at the top of the Channels palette to make it active, and deselect the new alpha channel. Choose the Eyedropper in the toolbox and click on the first letter A you made to make its color your foreground color. Choose Edit > Fill and pick Foreground from the menu.
     Now, compare the two letters. They are identical.

Just in case you’re still doubtful, here is a magnified view of the bottom of the two A’s shown above.

Ah, but this magnified view brings us to the next part of what masks and selections do. Notice that some of the pixels are neither red nor white, but are various shades of pink. Anti-aliasing was checked in the Type Mask tool’s options bar when the letters were made. This is an image effect. It is a form of instruction. It says “add some intermediate shades just here, and here”. In fact, it is even more precise than that. It says “at this pixel, add this exact shade, and at this one, add this particular shade.” Masks and selections are pixel maps telling Photoshop, How Much? for every single pixel.
     Let’s look at a close up of the mask at the enclosed part of the letter A, shown below.

I’ve made a screen shot of an extremely magnified portion of the bottom right corner of the triangle, and numbered all the shades of gray. Remember, this is the mask, not the red image.

If you choose the Eyedropper, and drag it over the shades of gray, the Info palette will show the following results:

1. Pure white
2. 
3. 
4. 
5. 
6. 
7. Pure black.

Note that you are seeing RGB values since this is a screen capture, not the original mask. The mask would show only K, or black percentages as masks are grayscale, not RGB.
      When that mask was loaded as a selection, and filled with a color, I got the result shown below.

As you can see, the mask has provided pixel by pixel instruction for how much the color you added should affect the existing, base image (in this case, white) pixel’s colors. More here, less there.

Let’s look at a photograph. I have added a color gradient to the base of the image to help you see what is happening when I show you individual color channels.

Older Tutorials
Quick Mask
Adjusting Contrast
Using Curves
Scanning Negatives
Rubberstamping
More Adjustments
Sharpening
Filters
Color

Here is what I got. As you can see, where the mask was lighter, the white fill color has been allowed to affect the base image to a greater degree. Where the mask was darker, it has allowed the white to affect the base image less.

A reading taken with the Eyedropper from exactly the center of the mask (not the image)

would show a reading in the Info palette like this:

Copyright © 2002 by Jay Arraich.
All rights reserved.
All photographs copyright ©2002 by Jay Arraich
jay@arraich.com