

In Adobe Photoshop, we can quickly convert a RGB image to true grayscale by clicking Image > Mode > Grayscale and the image is converted to black and white. However, this method is destructive and does not allow us to manage tone by color. The methods I devised and discuss here overcome these disadvantages while maintaining grayscale as the desired tonality.
The three basic Photoshop controls for tonal adjustments are Hue/Saturation,
Levels and Curves. Generally, it is best to use Levels or Curves for tonal
adjustments.
We can use all three Photoshop controls to adjust tone by
color ranges. We can use Levels and Curves to adjust tone by
color channel.
This section of the web site does not go into the details of how to use
Hue/Saturation, Levels or Curves to make tonal adjustments. That
information is covered on the
Hue/Saturation page, the
Levels page
and the
Curves page.
For in-depth information, read the
Using Levels and Curves with Color Ranges page.
The first example uses Curves for tonal adjustments by color. The second example uses Levels.
The GCC method preserves Grayscale luminance, uses a Channel Mixer adjustment
layer for monochrome conversion, and a Curves adjustment layer for tone
management by color. The step-by-step instructions to use this method are
as follows.
After all global and local changes have been made …
At this point, we now have a copy of the original image with all global and
local changes applied. This new document will become the final black and
white RGB image. The reason it has to be copied is to avoid double
applying the global and local changes when we apply the grayscale 'filter'.
Figure 1. Using A Grayscale Image As A Luminosity Filter
At this point, the only open document should be Grayscale RGB, and its
tonality matches that of a true grayscale image even though it is still in
color.
Figure 2. Channel Mixer Settings
We will now create the necessary adjustment layer to convert the image to
monochrome and allow us to control tone by color.
Above the image and above the Grayscale layer...
Tip
On the Downloads page is a Photoshop Action that will perform steps 1 through 23 (except step 3) for the photographer.
Figure 3. Curves Dialog Box
Steps 16 through 23 convert the color RGB image to a black and white RGB
image. The percentages in steps 17 through 19 are critical for preserving
the luminosity from the Grayscale layer. Do not change these percentages.
Instead, to change an image's tone, use the Levels or Curves adjustments.
Steps 27 and 32 are critical to ensure the changes we make only affect tone, not
color. Steps 24 through 29, or steps 30 through 35, are where we manage an
image's tone by color.
Figure 4. GCC Layers
Figure 4 is an example of what the Layers panel for the GCC method should look like if done properly. Note the blending mode of the Curves adjustment layer.
I realize this appears to be rather involved. Especially with creating two
copies of the image. However, if the grayscale layer were copied back to
the original image, then any global and local changes will be applied twice.
Once as part of the original image and again when the image was flattened and
then copied back to the original.
By changing the blending mode of the Grayscale layer, we turned it into a
luminosity 'filter'. Basically, it allows the color to pass through but
replaces the luminance in the underlying image with grayscale luminance.
Since color still passes through, we can manage tone by color while retaining
grayscale luminance. The Channel Mixer layer does not replace the
grayscale luminance with its own. Rather it preserves the grayscale
luminance.
The advantages of the GCC method are many. For those who prefer grayscale
tone, we preserve grayscale tonality while maintaining the ability to adjust
tone by color. The method is
non-destructive and since we are using Curves and/or Levels for tonal
adjustments, it is predictable.
There are three main disadvantages. First, we end up with two images.
Our master image and a black and white RGB version of it. Second, if we
decide to go back and change any of the global and/or local adjustments on the
master image, we will have to recreate the black and white RGB image from the
beginning. So it is best to be sure we are done with the color version of
the image before working on the black and white version. Third, because of
the number of steps involved, it is not easily intuitive.
The second method to be discussed is the GHL method. The GHL method preserves Grayscale luminance, uses a Hue/Saturation adjustment layer for monochrome conversion, and a Levels adjustment layer for tone management by color. The step-by-step instructions to use this method are as follows.
Figure 5. Hue/Saturation to Remove Color
Above both the image and Grayscale layers...
Tip
On the Downloads page is a Photoshop Action that will perform this conversion for the photographer.
Figure 6. Levels Dialog Box
Figure 7. GHL Layers
Steps 2 through 5 convert the color RGB image to a black and white RGB image.
Step 5 is critical to limit the desaturation adjustment to color, not tone.
If we do not change the blending mode to Color, Photoshop will calculate the
resulting tone using a different algorithm. (More information about this
can be found on the
Customary Method - Desaturation page.) By changing the blending mode
to Color, we drastically change the way the Hue/Saturation adjustment layer
desaturates an image. The Color blending mode causes Photoshop to
remove the color with affecting luminance. Steps 5 through 10, or steps 11
through 16, are where we manage an image's tone by color.
The adjustment layers below the Hue/Saturation adjustment layer allow us to
adjust tone by color range or color channel. Since the image is black and
white 'above' the Hue/Saturation adjustment layer, we cannot use color to target
our tonal adjustments above this layer. Therefore, any tonal adjustments
we need to make that are not based on color can be made above the Hue/Saturation
adjustment layer.
Figure 7 is an example of what the Layers panel for the GHL method should look
like if done properly. Note the blending mode for the Hue/Saturation
adjustment layer.
The advantages of the GHL method are many. Grayscale luminance is
accurately retained. We still have control over tonality by color.
The method is non-destructive and predictable.
Figures 8A and 8B show before and after images using either the GCx or GHx methods.
|
|
Tip
Once an image has been converted using GCx or GHx, we
can go to the Photoshop Layers panel and click the eye icon
on and off on the Grayscale layer. When the eye icon is on, the base
luminance is grayscale. Off, it is
measured luminance.
The GCx and the GHx methods give almost identical results. In my tests,
converting to monochrome using the Hue/Saturation adjustment layer (GHx), when
used with the correct blending mode of Color, gives slightly more accurate
results than Channel Mixer (GCx). However, the difference is imperceptible
(around 1/25 of a stop). Therefore, we can use either one. The
critical steps are getting the adjustment layers in the proper order and getting
the blending modes correct.
The first method above uses a Curves adjustment layer to manage tones by color. The second method uses a Levels adjustment layer. Which is better? The one you are more comfortable with. Levels has the advantage of giving us a histogram so we can readily compare the image's tonal range to the digital tonal range. Curves has the advantage of giving us more control along the entire tonal range. Levels and Curves can easily be used to make adjustments by color channel. With a little bit of additional work, they can also be used with color ranges. Hue/Saturation gives us the least amount of control over the tonal range, but it is the easiest control to use when working with color ranges. Hue/Saturation cannot be used with color channels.
Copyright © 2004-2009 Thomas Zuber. All rights reserved. Accessibility | Terms of Use | Contact Us | Legal Information | Privacy Policy | W3C Validation Service