The Hidden Power of Photoshop CS: Chapter 2: Color Separations. Pt. 1. By Sybex | 2 | WebReference

The Hidden Power of Photoshop CS: Chapter 2: Color Separations. Pt. 1. By Sybex | 2

The Hidden Power of Photoshop CS: Chapter 2: Color Separations. Pt. 1

Extracting the Brightness

Another means of changing an image to grayscale from RGB is to extract the image tone. You can do this in several ways with somewhat different results. Essentially you want to display the image brightness, based on grayscale as measured by luminosity in Lab mode or Brightness/Lightness in HSB (Hue, Saturation and Brightness) or HSL (Hue, Saturation and Luminance). Two easier methods are to extract directly from RGB or extract the luminosity after converting to Lab mode.

Use the Pumpkin.psd image from the CD to make the following conversions (see Figure 2.1 and the color section).

Extracting Grayscale from RGB Using Channel Properties

You can extract the tone right from an RGB image without the averaging that desaturation imposes. One feature of channels is that they can be loaded as selections based on their content: Just hold down the F/Ctrl key and click on the channel you want to load. Brighter areas of the channel will be selected while darker areas will not. Loading a channel as a selection can be used to create a tone extraction as in the following steps:

1. Open Pumpkin.psd and create a new layer at the top of the layer stack. Name it Brightness From Channels.

2. Fill with white, and shut off the visibility for the layer.

3. Hold down the z/Ctrl key and click on the RGB composite channel in the Channels palette. This will load the channel as a selection based on its brightness.

4. Activate the Brightness From Channels layer created in steps 1 and 2.

5. Fill with black. The image will appear as a negative.

6. Deselect and invert the layer (Image ‘ Adjustments ‘ Invert).

This result is the image brightness based on RGB tone—a result of loading the channel as a selection.

Figure 2.1 This pumpkin has some pretty vivid coloring (see the color section), which will translate in different ways to grayscale.

Extracting Brightness from RGB Using Layer Properties

Another way to accomplish about the same thing is to extract the brightness using layer properties:

1. Using the same image, shut off the view for the Brightness From Channels layer and create a new layer at the top of the layer stack; then fill with white and shut off the layer view by clicking the Adobe eye icon.

2. Create another new layer, then press z+Option+Shift+E/Ctrl+Alt+Shift+E to merge visible to the top layer.

3. Change the layer mode to Luminosity.

4. Turn on the view for the layer created in step 1 (click the Adobe eye, but don’t activate the layer).

5. Merge the two layers created in steps 1 and 2.

6. Name the layer Brightness From Layers.

If you compare the Brightness From Channels and Brightness From Layers layers, you may see a slight difference. To compare visually, toggle the view of the Brightness From Layers layer. To make a digital comparison, you can set the layer to Difference mode. Difference Mode will reveal any difference as lighter areas against black. These may be hard to see with the naked eye, but you can note the difference by viewing a histogram.

Converting to Grayscale from Lab Color

The Lab color model breaks down color using channels that represent brightness and color balance. The Lightness channel represents the tone, while the a and b channels represent the color balance between redgreen and blueyellow, respectively. The result is somewhat different than when you extract tone from RGB because of the difference in
the color modes themselves. RGB works only with mixing color; there is no pure tone representation. Lab’s Lightness channel directly represents image tone.

This extraction can also be done in several ways. All you are looking to do is convert to Lab mode and copy out the Lightness channel to an RGB image.

EXTRACTING THE LUMINOSITY FROM A LAB IMAGE BY DUPLICATING LAYERS

First, let’s see the steps involved in extracting the luminosity from a Lab image by duplicating layers:
1. Using the Pumpkin.psd image, shut off the view for the Brightness From Channels and Brightness From Layers layers.

2. Click the Background in the Layers palette to activate it and copy it to a new image. (Use Duplicate Layer from the Layers palette pop-up menu and select New as the target.)

3. Change the mode of the new image to Lab. To do this, select Image ‘ Mode ‘

Lab Color.

4. Click on the Lightness layer in the channels.

5 Change to Grayscale mode and discard other channels when offered the option by clicking OK in the dialog box that appears. Since the appearance is already grayscale, you should see no change in the image.

6. Choose Duplicate Layer from the Layers palette and select Pumpkin.psd as the target. The grayscale background of the converted image will appear in the Pumpkin.psd document as Background Copy.

7. Change the name of the Background Copy layer to Luminosity.

EXTRACTING THE LUMINOSITY FROM A LAB IMAGE WITH APPLY IMAGE

A second way to move the extracted luminosity from a Lab image is to the use Apply Image command. The advantage of this approach is fewer steps, numerous options for application and masking.

1. Be sure all other layers you added are off and add a new layer to the top of the layer stack in Pumpkin.psd. Call it Applied Image.

2. Duplicate the background from Pumpkin.psd to a new image.

3. Change the duplicate image to Lab mode.

4. Activate the original image, Pumpkin.psd.

5. Choose Image ‘ Apply Image and use the following settings (see Figure 2.2):

Figure 2.2 Apply Image is a powerful tool that can be used for quick calculations between open images. These settings will complete the duplication between images for you in the current example.

Once you complete step 4, the Luminosity channel from the duplicate will be copied to the top of the layer stack in the original image. You can achieve nearly the same result by changing the image to Lab Color and then Grayscale mode and back to RGB. The result is not exactly the same as direct Lab conversions, because the conversion process includes some extra steps. The differences can be traced back to your color-management settings.

Just to prove a point, try this:

1. Shut off the visibility of all the layers except the Background, then create a new layer at the top of the layer stack in Pumpkin.psd. Name it Desaturate.

2. Press z+Option+Shift+E/Ctrl+Alt+Shift+E to merge visible to the layer created in step 1.

3. Desaturate using the Desaturate feature (Image ‘ Adjustments‘ Desaturate), or press z/Ctrl+Shift+U).

4. Compare the grayscale results by Option/Alt clicking on the visibility icon for the Luminosity, Brightness From Layers, Brightness From Channels and Desaturate layers. This will turn on the view for the view for the layer you click and shut off the others.

In step 4, the differences between each layer should be apparent on your screen. You should see a difference in each of these results (see Figure 2.3), but which is the best may not be clear. This will happen no matter what image you use. For some images one conversion will look best, and for another a different conversion will work well. These simple examples represent only the start of the subtle differences you’ll note in conversions. You

Figure 2.3 The colorful pumpkin (see the color section) can easily break out into different grayscale representations depending on how it is converted.

can achieve a very different result by working with tone from separations into RGB and CMYK, and by combining results from different channels and separations. The point is that there is no best process or best conversion approach to getting black-and-white from color. You have to know many methods, and you have to try to visualize results to get them. Sometimes the easy route pays off, and sometimes you need to go a long way to make a great color image look decent in black-and-white.

Created: March 27, 2003
Revised: March 1, 2004

URL: URL: http://webreference.com/graphics/ps1/1