Image components are separations of an image into distinct color or tone parts. There are many ways to separate images into other types of components, including color components of light (red, green, and blue) and ink (cyan, yellow, magenta, and black). Separating images into components can offer advantages in making corrections, such as creating masks, setting up calculations, and converting images to black and white.

Separations provide an essential understanding of how images are composed, stored, and viewed. Being able to work with color components directly as separations is nearly the exclusive reason for channels—to which Adobe has dedicated an entire palette in Photoshop. When the Photoshop user learns to look at channels as component parts of images in the same context as layers (rather than as part of a separate palette), he or she gains many times the potential flexibility. Working with channels/components in layers leads to a better understanding of how they fit into images and how they can be used directly in corrections rather than reaching to a separate part of the interface.

One of my favorite digital lessons is learned from taking a set of blackand- white images created before there was color fi lm and making a color representation of the image. A special case is the photos of one Russian photographer, Sergei Mikhailovich Prokudin-Gorskii. You can find digitized images in several libraries online, here and here.

Using a special camera that he designed, Prokudin-Gorskii captured images on glass plates three at a time (it is said in rapid succession, rather than all at once). During the capture, color filters separated red, green, and blue color components to different areas of the plate. The result was a single plate with black-and-white representations of the image’s core light components (see below).

The solution still offered only black-and-white representations of the RGB channels, was a bit awkward, and required a customized projector to reproduce the color. But really these first color captures mimic what digital cameras do even today, separating and storing color in red, green, and blue light components. About 100 years after they were taken, we can treat Prokudin-Gorskii’s images as components of an image and put them back together using Photoshop to re-create their full-color representations.

A scan of an original Prokudin-Gorskii "color" plate taken between 1907 and 1915, 20 years before Kodachrome... the first color film. Stacked here from the top down are the blue, green, and red color components.

The solution still off ered only black-and-white representations of the RGB channels, was a bit awkward, and required a customized projector to reproduce the color. But really these first color captures mimic what digital cameras do even today, separating and storing color in red, green, and blue light components. About 100 years after they were taken, we can treat Prokudin-Gorskii’s images as components of an image and put them back together using Photoshop to re-create their full-color representations.

Working with separations provides some valuable background for what we’ve already been doing in correcting for different components of light with Levels. It also opens doors to additional techniques for working with color and black-and-white images.

The concept of RGB and the idea that an entire world of color can be stored in combinations of three colors really doesn’t seem plausible until you see it at work. That is, the 16 million color variations in 8 bits per channel and 35 billion in 16 bits per channel are all produced from the capture of red, green, and blue core components.

In the following short example, we’ll look at putting together a Prokudin- Gorskii image from his original black-and-white captures.

Try It Now

1. Download the components.zip archive, extract the files it contains and open the three Samples_15 images. They are named Sample_15_red.psd, Sample_15_green.psd, and Sample_15_blue.psd. Be sure to keep them in order, or your result will not turn out correctly (see below). (If you’d like an extra challenge, there is a Sample_15_RGB.psd in the archive as well. This image has a scan of the complete glass plate. You can work from that if you’d like, but be forewarned that there are issues of alignment and distortion that have mostly been addressed in the cropped version provided.)

The three sample images represent the red, green, and blue components.

2. Activate the Sample_15_blue.psd image, then press Command+A/Ctrl+A (Mac/PC) to select the entire image, then press Command+C/Ctrl+C to copy. Doing this stores a copy of the image as well as the image dimensions, which helps automate the next step.

3. Create a new image (File > New), name the file Prokudin-Gorskii Composite, as shown below, and be sure to change the Color Mode to RGB Color (it will initially be Grayscale). We will use this new image to assemble a color image from the components in the other three images.

Your New dialog should look very much like this when opened, as Photoshop will have automatically defined the new image size from the image information on your clipboard.

4. Create a new layer, call it Compositing Screen, and fill it with black (Edit > Fill and set the Use content to Black in the drop-down list). This will act as a projection screen for the image components.

5. Press Command+V/Ctrl+V to paste the content of the clipboard to the image. Name the resulting layer Blue.

After pasting the initial component, you will have three layers: Background, Compositing Screen, and Blue.

6. Arrange the Prokudin-Gorskii Composite image and the Sample_15_ green.psd so you can see both on your monitor, choose the Move tool (press V), hold the Shift key, and click and drag the Sample_15_green.psd image into the Prokudin-Gorskii Composite file. Release the mouse button and then the Shift key. Name the new layer in the Composite image Green.

7. Arrange the Prokudin-Gorskii Composite image and the Sample_15_ red.psd image so you can see both on screen, choose the Move tool (press V), hold the Shift key, and click and drag the Sample_15_red. psd image into the Prokudin-Gorskii Composite file. Release the mouse button and then the Shift key. Name the new layer in the Composite image Red. The result of all this clicking and dragging should look like the illustration below.

The resulting layers from the top down should be Red, Green, Blue, Compositing Screen, and Background.

8. Close the Sample_15_red.psd, Sample_15_green.psd, and Sample_ 15_blue.psd images, leaving only the Prokudin-Gorskii Composite image open.

9. Shut off the views for the Blue and Green layers so you are viewing only the Red layer. It will still appear in black and white at this point.

10. Double-click the thumbnail for the Red layer in the Layers palette. The Layer Style dialog will open.

11. Uncheck the Green and Blue checkboxes for Channels under Advanced Blending (see below). The image will turn red, showing a view of how the red light component in the image looks when isolated.

Unchecking the checkboxes for Green and Blue makes the red layer act only on the Red channel. The result is that you see the red light component in red.

12. Shut off the Visibility toggle for the Red layer and toggle the view for the Green layer so it is visible again. The image will appear as a grayscale representation of the Green channel.

13. Double-click the thumbnail for the Green layer in the Layers palette for the Prokudin-Gorskii Composite image. The Layer Style dialog will appear.

14. Uncheck the Red and Blue checkboxes for Channels under Advanced Blending. The image will show the green component channel in green (see below).

Unchecking the checkboxes for Red and Blue makes the green layer act only on the Green channel. The result is that you see the green light component in green.

15. Shut off the Visibility toggle for the Green layer, and toggle the view for the Blue layer so it is visible again. The image will appear as a grayscale representation of the Blue channel.

16. Double-click the thumbnail for the Blue layer in the Layers palette for the Prokudin-Gorskii Composite image. The Layer Style dialog will appear.

17. Uncheck the Red and Green checkboxes for Channels under Advanced Blending. As you might expect, the image will show the blue component of the image in blue (see below).

Unchecking the checkboxes for Green and Red makes the blue layer act only on the Blue channel. The result is that you see the blue light component in blue.

18. Now turn on the Visibility toggles for the Green and Red layers. You will see a full-color composite of the image, though there is no color at all in any of the layers in the Layers palette (below).

The result has some issues as far as needing color correction, dynamic range enhancement, and some obvious cleanup, but the fact is that you have just created color from images taken about 100 years ago, 20 years before there was color film.

Use of those little checkboxes on the Advanced Blending panel of the Layer Style dialog demonstrates several important light theory concepts. Layer Styles target the content of the layers to make each act like a specific light component, which projects on the dark screen. When all three of the light components are switched on, the red, green, and blue components combine to re-create the full-color image from the three separate black-and-white components. This is what your monitor projects, what your camera captures, and what Photoshop re-creates: red, green, and blue components are assembled to make a color image. The example shows that colored light works when stored as grayscale measurements and is a sort of demonstration of what Prokudin-Gorskii needed to do to re-create the color he captured: project the filtered images onto a screen.