The Digital Darkroom in Adobe Photoshop CS4
Adjusting Images Using Blending Modes
Applying Levels, Curves, Hue/Saturation, and the other adjustments as adjustment layers offers tremendous flexibility and power, but they aren’t always the quickest or easiest ways to fix your images. In some cases, you may find it easier and faster to do the heavy lifting with blending modes, and save Curves for fine-tuning.
Using Layer-Blending Modes
A blending mode takes the color values of two layers and combines them using a formula, so it’s more sophisticated than just changing opacity. Some blending modes happen to have effects that conveniently help you with basic tone and color edits, and all you have to do is copy a layer on top of the original and apply a blending mode to the copy. You can achieve the same result by adding an adjustment layer and then change only its blending mode. That’s faster and easier than making a duplicate of an image layer, and it saves disk space. If the result of the blending mode isn’t quite perfect, just edit the adjustment layer (we typically use Curves) to cover the rest of the distance. When blending modes and adjustment layers join forces, their powers multiply many times over. If the following blending mode descriptions make your eyes glaze over, hang on—examples are coming up.
How Blending Modes Think. The biggest question anyone has when using blending modes is, “How do I know which blending mode will do what I want?” To find out, ask yourself the following questions:
- Do I want to alter the light areas of the underlying layers, the dark areas, or both? Many blending modes have a neutral color that doesn’t change the layers under it. For example, for the Lighten, Screen, Color Dodge, and Linear Dodge blending modes, black pixels on the applied layer don’t alter the same pixels on the underlying layer, and the farther an applied pixel is from black (that is, the lighter it is), the more it affects the underlying layer.
- For other blending modes, white or 50 percent gray is the neutral color. When 50 percent gray is the neutral color, that means 50 percent gray pixels on the applied layer don’t affect underlying pixels at all, and the farther an applied pixel is from 50 percent gray (that is, the closer it is to black or white), the more it affects the underlying layer.
- How much additional contrast do I want? Part of the reason there are so many blending modes is that quite a few of them are merely variations of other blending modes that produce more or less contrast. For example, Hard Light is a higher-contrast version of Soft Light. Certain blending modes, such as Difference, take contrast to the extreme, inverting (creating the negative of) the original layer color. The blending modes that produce lower contrast tend to compare an applied pixel to an underlying pixel and simply keep one or the other, while the blending modes that produce higher contrast tend to mathematically amplify the differences between the applied and underlying layers.
- What do I want to change? Most blending modes affect any underlying pixel (other than those in the blending mode’s neutral color, if it has one), while some blending modes affect only color or tone.
The blending modes are arranged in logical groups, according to the way they answer those three questions.
The Independent Modes. Normal and Dissolve both replace the underlying pixels with the pixels of the applied layer when the layer is at 100 percent opacity. At lower opacities, Normal blends the overlying pixels with the underlying ones according to the layer Opacity value, while Dissolve replaces pixels randomly (see Figure 8-14).
Figure 8-14 Blending mode examples
The Darken Modes. The neutral color for the Darken modes is white. White pixels on a layer set to a Darken mode leave the underlying pixels unchanged. Nonwhite pixels darken the result by varying amounts, depending on each blending mode’s math and the difference in value between the applied and underlying pixels (see Figure 8-15).
Figure 8-15 The Darken modes
The Lighten Modes. The Lighten modes are the inverse of the Darken modes. The neutral color for the Lighten modes is black—black pixels on a layer set to a Lighten mode leave the underlying pixels unchanged. Nonblack pixels darken the result by varying amounts, depending on each blending mode’s math and the difference in value between the applied and underlying pixels (see Figure 8-16).
Figure 8-16 The Lighten modes
The Contrast Modes. These modes combine corresponding Darken and Lighten modes. The neutral color for the Contrast modes is 50 percent gray—50 percent gray pixels on a layer set to a Contrast mode leave the underlying pixels unchanged. Lighter pixels lighten the result and darker pixels darken the result; the amount depends on the blending mode and the value difference between the applied and underlying pixels (see Figure 8-17).
Figure 8-17 The Contrast modes
The odd man out is the Hard Mix blend, which has no neutral color but doesn’t fit anywhere else either. It reduces the image to eight colors—red, cyan, green, magenta, blue, yellow, white, and black—based on the mix of the underlying and blend colors, with a strength related to 50 percent gray.
The Comparative Modes. The neutral color for the Comparative modes is black. The Comparative modes look at each channel and subtract the underlying color from the overlying color or the overlying color from the underlying color, whichever returns a result with higher brightness. Blending with white inverts the underlying color values (see Figure 8-18).
Figure 8-18 The Comparative modes
The HSL Modes. Blending modes in other groups generally operate on overall tone and color values. The members of the HSL group work with hue, saturation, and luminosity (HSL) instead (see Figure 8-19).
- Hue. This blending mode creates a result color with the brightness and saturation of the underlying color and the hue of the overlying color.
- Saturation. This mode creates a result color with the brightness and hue of the underlying color and the saturation of the overlying color.
- Color. This mode creates a result color with the luminosity of the underlying color and the hue and saturation of the overlying color.
- Luminosity. This is the inverse of the Color blending mode. It creates a result color with the hue and saturation of the underlying color and the luminosity of the overlying color.
Figure 8-19 The HSL modes
Layer Blending in Practice
Despite the mind-bending variety of blending modes, for the purposes of tonal and color correction we tend to use just a few of them. We use Multiply to build density, Screen to reduce it, Soft Light and Hard Light to increase contrast, Color to change color balance without affecting luminosity, and Luminosity to sharpen images without introducing color fringes (we discuss sharpening layers in more detail in Chapter 10, “Sharpness, Detail, and Noise Reduction”).
The practical example that follows doesn’t pretend to exhaust the power of blending modes, but we hope it fires your imagination and gives you alternative ways of approaching problems. Blending modes are so quick and easy for basic adjustments that the image practically edits itself.
Building Density Using Multiply. Multiply mode always creates a result that’s darker than both the layer you apply it to and the layer behind that. If you’ve worked in a darkroom, it’s like sandwiching two negatives in an enlarger. Mathematically, Multiply takes two values, multiplies them by each other, and divides by 255.
If a pixel is black in the base image, the result after applying an adjustment layer with Multiply is also black. If a pixel is white in the base image, the adjustment layer has no effect (because white is the neutral color for Multiply). We use Multiply with Curves adjustment layers to build density, particularly in the highlights and midtones of washed-out images. In Figure 8-20, we use Multiply to knock down the near-blown-out sky highlights.
Figure 8-20 Using Multiply to rescue highlights that are almost blown out to white
Opening Shadows Using Screen. Screen is literally the inverse of Multiply. A popular analogy is that Screen is like projecting two slides on the same screen. The result is always lighter than either of the two sources.
If a pixel is white in the base image, the result is white, and if it’s black in the base image, the result is also black (black is the neutral color for Screen). Intermediate tones get lighter. We use Screen to open up dark shadows.
If you want to know the math, Photoshop inverts the two numbers (subtracts them from 255) before doing a Multiply calculation (multiplies them by each other and divides by 255); then the program subtracts the result from 255.
In Figure 8-21 we open up the shadows of the image we just worked on. To save time, we use an inverted copy of the mask we’ve already made.
Figure 8-21 Using Screen to open up shadows
Adding Contrast Using Hard Light. We use Soft Light, Hard Light, and Overlay to build contrast (since the overlying and underlying pixels are identical, Hard Light and Overlay produce exactly the same result). We use Soft Light for smaller contrast boosts and Hard Light or Overlay for stronger ones. All three blending modes preserve white, black, and 50 percent gray while lightening pixels lighter than 50 percent gray and darkening those that are darker.
In Figure 8-22 we show an example of using Hard Light. Not surprisingly, it gave us a rather extreme result. We chose to back it off by reducing the opacity of the Curves adjustment layer, but another alternative would have been to try another contrast blending mode, such as Soft Light.
Figure 8-22 Increasing contrast using Hard Light
Adjusting Color Balance Using Color. While we sometimes use Photo Filter adjustment layers for warming and cooling effects, we find that a Solid Color layer set to the Color blending mode and low opacity offers more control. Tweaking the color by double-clicking the layer thumbnail in the Layer panel is faster than tunneling through the Photo Filter dialog.
We start out by creating a Solid Color layer of approximately the color we want, then we reduce the opacity, typically to around 10 to 20 percent. We then fine-tune the color to get the final result, usually tweaking the Hue and Saturation fields of the Color Picker by placing the cursor in them and pressing the up and down arrows on the keyboard. It’s very similar to the way we edited the target midtone value of the Auto Color Correction settings in Figure 7-10 in Chapter 7.