Last week’s tutorial about using the ExpoDisc to capture good skintones on camera generated lots of interest. And convinced me that it’s time to revisit the technique behind editing skintones in Photoshop Elements.
The most important tool Elements gives us to correct skintones is the Info panel, which lets us view color measurements. To view your Info panel, go to the Window Menu and click on Info.
It looks like the screen shot below, when visible. If the top left quadrant of the Info panel doesn’t display “RGB Color”, select it by clicking on the arrow next to the name of that quadrant. As you move your cursor over your image, the Red, Green & Blue fields will reflect measurements of the color components of the pixels you are hovering over.
For best results when measuring colors, select your Color Picker tool (the Eyedropper), set the sample size to a 5×5 average, and set Sample to All Layers when measuring.
Go ahead and give this a try, now that you’ve got it all set up. Move your cursor around and examine the measurements that appear in the Info panel.
Are you ready to learn what those numbers can tell you?
Let’s take this photo as an example. Start on a photo whose overall white balance is good.
The measurements for the pixels at the end of each arrow are in this table:
|Point 1||Red: 231|
|Skin tone & exposure pretty good|
|Point 2||Red: 170|
|Skin underexposed in shadows with strong orange cast|
|Point 3||Red: 221|
|Skin exposed well with orange reflecting up from dress|
We can make these observations about the numbers:
- Red is always greater than green, which is always greater than blue. This is good for skintones and you will see this pattern represented in nearly every skin color of every ethnicity.
- The higher the numbers, the lighter the colors. See how low the values are on point 2, in the shadows? RGB numbers run on a scale from 0-255, with 0 being the lowest and 255 the highest.
To repeat some info from last week’s article, you will see neutral colors whenever red=green=blue. Here are examples of neutral colors:
- Black – Red 0, Green 0, Blue 0
- White – Red 255, Green 255, Blue 255
- Dark gray – Red 50, Green 50, Blue 50
- Light Gray – Red 175, Green 175, Blue 175
Colors represented by unequal amounts of red, green and blue are, well, colors. They aren’t neutral shades of gray. To understand what the measurements mean, you first have to understand that each color, as measured in Elements, has an opposite.
For example, blue and yellow are opposites and are inverse to each other. At Point 2 above, Blue measures 79. As its inverse, Yellow would be 176 (or 255-79). So the lower Blue is, the higher Yellow is. And the lower Yellow is, the higher Blue is. This works for these 3 color pairs:
Now that this makes sense (it does make sense, right?), we can make another observation about the numbers in the screen shot above. Blue is much lower at Points 2 and 3 than at 1. So yellow is much higher. And given that Red is the largest number at each of these points, when we combine the Red with the larger than normal amount of Yellow, we get Orange. Which is exactly what our eyes see.
In this photo, the orange color cast is quite obvious and the RGB numbers merely confirm what our eyes see. Using the Info panel’s numbers is especially helpful for photos where you know the skintones are off, but aren’t quite sure why. Unfortunately, problems aren’t always as obvious as they are in this photo!
Once you’ve used the numbers to identify what needs to be fixed, use one or more Levels adjustment layers to fix the affected areas. For the purposes of this discussion, I’m going to assume that Point 1 needs no edits.
To correct the area around Point 2, I need to lighten and remove the excess Yellow. Using a Levels adjustment layer, I move the middle slider of the RGB channel to the left. Moving this slider to the left brightens a photo, and moving it to the right darkens a photos.
Next, on the Blue channel, I move the middle slider to the left to increase Blue and decrease Yellow. When adjusting the individual color channels in Levels, moving the middle slider always increases the color that the channel is named after (Red, Green or Blue). Moving the middle slider to the right always increases that color’s opposite. These are my Levels adjustments for Point 2.
I tweak both sliders until I am satisfied with the appearance of the dark, shadowed, orange-y area of her skin and finish by using the layer mask to constrain the edits just to this area.
For Point 3, the exposure is pretty good, so I only need to correct the orange color cast. I do this on a 2nd Levels adjustment layer, moving the middle slider of the Blue channel to the left. And I finish by constraining the effects of this layer with a layer mask also.
After these edits, you can see the following changes in my numbers:
|Point 1||Red: 231|
|Point 2||Red: 170|
|Point 3||Red: 221|
And here is a Before and After on the image. Note that the shadows are lighter in the After and the orange tones are minimized on the bottom half of her face and her neck. I could do a lot more to tweak the final photo, but this tutorial is plenty long enough as it is!
When editing your own photos, start by reading the color measurements at several points of your image. If you want to lighten darker parts of your image, start by increasing the middle slider on the RGB color channel of a Levels layer. Do this selectively, using the layer mask, to equalize lightness around the subject.
To correct the color channels, it will take experimenting and practicing. If blue happens to be larger than green, add green/reduce magenta and/or decrease blue/add yellow with the appropriate color channels. After that, explore the R to G to B proportions of various skin points, and try to equalize those proportions throughout the image. For example, if green is 15 higher than blue at a good point, try to get the same 15 point spread throughout the image.
Do you get the feeling that this iceberg is much huger than the tip you just explored? Oh yes, it is. But I think this is plenty for one article!