The Adjustments, Listed

The majority of color and tonal adjustments can be applied nondestructively as adjustment layers or Smart Filters (applied to Smart Objects). The remaining few can be applied only directly to pixels, and thus care should be taken if you might ever wish to reverse their effect later. In the explanations below, where possible, I have used the icons found in the Adjustments panel for easier identification. Also, I’ve divided adjustments into two main categories: productions and creative, since some lend themselves to practical use more than others. Of course, many can be used in either way.

To see the full list of Photoshop adjustments and to apply them directly or as Smart Filters, go to Image > Adjustments. The most important of those can also be applied by clicking one of the following icons in the Adjustments panel:

When using the Adjustments panel to create an adjustment layer, the Properties panel will show the controls for the adjustment as well as the following common elements:

Production Adjustments

The boundary between “Production” and “Creative” adjustments is a amorphous one. You should feel very free to use those listed in each category however you choose. But note, the first set is used in practical, production workflows far more frequently than the latter.

The Properties panel is where we configure adjustment layers, and we’ll focus on those. I’ll deal with the few adjustments that can be applied only destructively at the end of this section.

Brightness/Contrast

This adjustment offers only two sliders, one for Brightness, which behaves like the Levels adjustment’s midtone slider, and one for Contrast, which approximates a Curves adjustment with either an S-shaped curve (for increased contrast) or an inverse S. See “Levels” (page 230) and “Curves” (page 233) for more on those more powerful adjustments.

The Auto button is the simplest way to adjust these two attributes simultaneously; it may at least offer a starting point for your adjustment. Once dismissed as horrible, Brightness/Contrast is now quite useful, though it offers little control. You may find that this tool works great if the color does not need to be adjusted or if you’d like to deal with tonal and color adjustments separately.

To ensure that you are not clipping highlights or shadows, keep an eye on the Histogram panel. See “Histograms” (page 226).

Levels

More precise than Brightness/Contrast but less so than Curves, Levels is a great tool for setting black and white points, making the overall image lighter or darker, or doing quick color corrections. More so than with other adjustments, using Levels feels like we’re directly interacting with an image’s histogram.

The objective for an image like this may be to bring out more shadow detail (in the sculpture) without sacrificing the highlights (sky) too much or muddying the blacks.

The illustrated adjustment lightens the image:

• Black remains black: Both input and output black levels are still set to 0.
• White point adjusted: An input level of 195 outputs to white (255).
• Midtones significantly adjusted: There are now 2.50 times as many original levels that are lighter than middle gray as there are those that are darker. In the Histogram panel, this has the effect of stretching the levels on the left and compressing those on the right, since the middle gray is still in the middle.

Clipping Preview

To monitor which and how many pixels are getting clipped while adjusting the black and white points, hold down option/Alt. As you option/Alt-drag the black point, the image will turn white except where shadow clipping is occurring. As you option/Alt-drag the white point, the image will turn black except where highlight clipping is occurring.

The previous adjustment significantly increases contrast:

• Black input at 39 means that all levels below (left of) that become 0, the output black level.
• White input at 113 means that all levels above (right of) that become white (255 output level).
• Midtones are still set at 1.
• Since there were few three-quarter tone pixels, the image has become a silhouette, with its histogram showing spikes at black and white, indicating massive shadow and highlight clipping.

You can perform this contrast enhancement on each channel too. Select the channel whose levels you wish to edit from the Channel menu in the Properties panel. Instead of white or black for a clipping warning, you’ll see the color of the channel you’re editing.

There are automatic-adjustment algorithms you can access by option/Alt-clicking on the Auto button. See “Auto Tone & Auto Color” (page 235) in the description of the Curves adjustment.

Curves

To refine tones with far more precision than Levels, use the Curves adjustment. Curves is commonly used to adjust contrast in the image and to make precise adjustments to individual tonal ranges in the image as a whole or on a per-channel basis. It does everything Levels does and much more. For that reason, you may wish to look over the description of “Levels” (page 230).

When you first choose a Curves adjustment, the “curve” is a straight line from the lower left to the upper right of the window in which you see a histogram and do your adjusting. As you tweak the curve, a faint baseline indicates where you started.

The curve is essentially a graph where we compare and affect tones coming into the adjustment (input levels) to the tones they become (output levels). By simply dragging the curve up or down, we create a point where we drag, making the image lighter or darker, respectively. If you prefer dragging up to make an image darker (as if you’re adding ink to the image rather than light), use the adjustment layer’s Properties panel to choose Curves Display Options.

In the Curves Display Options dialog, you can make the graph a more granular 10 x 10 grid, or choose to hide the Baseline, Histogram, or the Intersection Lines you see when you drag a point on the curve. By default, black is at the lower left and white at the upper right. In the screenshot of the Properties panel on the previous page, the active point (B) is being dragged slightly above the baseline position. Its Input reads 98 and its Ouput is 104, a slight lightening, but the curve is still close to the baseline near the midtones, leaving them almost unaffected by this adjustment.

The other point (A) is well above where it started, creating a dramatic lightening of the shadows, represented by the large spike on the left side of the histogram. Since the curve is steep through that part of the histogram, contrast has been increased in that part of the tonal range. In fact, that is what I was attempting to do through both large humps in the histogram. However, when we add contrast to one part of the tonal range, we’re borrowing it from adjacent tones. Luckily, in this case, there were few if any pixels in several parts of the histogram (anything lighter than 204, and a sizable range between the shadows and midtones). By allowing those to become flatter, I could give contrast to the parts that are represented.

So I dragged the white-point slider, moving the white point (C) to 204 as Input, so that tone and all lighter tones would Output to 255 (white). I created and then lowered the point near the midtones (B) to create a steeper curve through the new highlights and to keep the image from becoming too light overall.

Summary: The parts of the histogram where the curve is steeper experience increased contrast. Where the curve is flatter, so is the image (“flat” = “lacking contrast”). The Curves adjustment allows us to lighten, darken, and affect contrast differently across the entire tonal range.

Auto Tone & Auto Color

Curves is much more effective than Levels, as it offers everything Levels does, including the automatic settings accessed by option/Alt-clicking on the Auto button. Doing so opens a dialog with four algorithms to choose from, as well as a few other refinements.

When you’ve used this dialog to apply an adjustment that seems to be a better default than the first algorithm (which is very likely), enable Save as defaults. Then the adjustment you used will be what’s applied when you simply click the Auto button in Levels or Curves.

Exposure

This adjustment is most effective and useful with 32 bit/channel HDR images, allowing you to see details in different parts of that potentially huge range of tones. It is not terribly useful for anything else. It should always be used as an adjustment layer so as not to lose HDR data.

When choosing this adjustment, the Properties panel will offer three sliders. The Exposure slider has the largest impact, but affects shadow areas minimally. The Offset slider affects shadows, often horrifically, but with minimal impact on the highlights. The Gamma slider affects contrast. When editing HDR images, you can access a version of the Exposure slider at any time at the bottom of the image window, as seen below.

I appreciate this feature when editing 32 bit/channel images, as it lets me see different parts of the tonal range, often hugely greater than that of my monitor, without doing any harm. I use these High Dynamic Range images most when I’m using Photoshop’s 3D features. There, they can be used as light sources that surround a 3D model. Interested? See “3D” (page 371).

Vibrance

Vibrance is a simple and effective choice for adjusting saturation. Its two sliders differ only in degree and scope. Both are restrained compared to the saturation slider in Hue/Saturation (despite having the same name).

In this adjustment, Vibrance increases saturation mostly in less-saturated areas to avoid color clipping. It also avoids affecting skin tones more than the Saturation sliders in either adjustment, especially the one in Hue/Saturation. In short, the Vibrance slider is my favorite way to adjust saturation safely.

Since these sliders affect different hues differently, reducing each to -100 produces different results too. Vibrance leaves some color lurking in what had been the most-saturated areas. The Saturation sliders in both the Vibrance and Hue/Saturation adjustments yield a colorless result, the former looking very similar to a conversion to grayscale mode, and the latter like a default Black & White adjustment.

Hue/Saturation

Despite its name, this adjustment actually affects three attributes: hue, saturation, and lightness (HSL). Unlike the Vibrance adjustment, you can choose to affect these three attributes differently for different hues. It often surprises users how precisely one can target hues for adjustment. You may have seen garments or other product photos in catalogs that differ from one another only in their color. Hue/Saturation very likely played a major roll in that.

When you first create a Hue/Saturation adjustment layer, the Properties panel shows that you’re adjusting all hues at once: the hue menu (Preset) is set to Master.

Note: There are two UI elements called the “hue slider!” The obvious one labeled Hue is above the Saturation slider and is used to alter hue in the image. The other appears between the color bars near the bottom of the panel when adjusting an individual hue and is used to fine-tune the hue being altered. For this conversation, I’ll refer to the upper one as the “Hue slider” and the lower one as the “hue slider.”

When you adjust the Hue, you offset colors’ positions on the color wheel. Thus, dragging the Hue slider all the way to the left or right changes the value in the field to +/- 180, as in 180º (the opposite hue). The color bars at the bottom of the panel offset to one another to indicate what each original hue (top color bar) has become (lower color bar). Double-click the word Hue to reset.

The Saturation slider can completely desaturate an image (-100%), much like converting to grayscale. It can also radically increase saturation to sickening levels (+100%). For more tolerable saturation adjustments, when that’s all you need, I recommend using the Vibrance adjustment. Double-click the word Saturation to reset.

The Lightness slider is rather blunt and horrible when the hue menu is set to Master. When adjusting a single hue, however, it is quite useful. Double-click the word Lightness to reset.

There are two ways to adjust one hue at a time. The more methodical way is to choose a hue like Reds or Blues from the hue menu. The multi-sectioned hue slider appears between the color bars at the bottom of the adjustment, centered under Photoshop’s definition of that hue. To ensure that Photoshop is adjusting the hue you want it to, select the first hue sampler (eyedropper), then click on pixels containing that hue. (The active part of the cursor is its lower left corner. For a precise crosshair cursor, enable your keyboard’s caps lock.) You’ll see that the hue slider has probably moved under the exact hues on which you clicked.

A faster way to get to the right hue is to use the On-image adjustment tool: . Choose it, then move the cursor over the image. It will look like the hue sampling eyedropper (or crosshair if caps-lock is on). When you click on a color, the hue menu will automatically change. Use the hue sampler dropper to center on the precise hue you want to adjust. If the hue slider is too close to one end, ⌘/Ctrl-drag the color bars to see all of it again.

Now, when you move the three adjustment sliders, only that hue will change. In the illustration on the previous page, we can see that not only reds are changing, but also oranges and yellows, since they are in the fall-off zone of the hue slider.

Color Balance

With this adjustment, you can adjust hues in different parts of the tonal range. It’s not as precise as Curves, but it is fast. I sometimes use it to remove blue from shadows by choosing Shadows from the Tone menu and then adjusting the balance toward Yellows from Blues.

Preserve Luminosity keeps the image from getting lighter or darker as you make adjustments. If those shadows are really blue, for example, they could get much darker if I remove blue without that option enabled.

Photo Filter

I love this adjustment! It’s simple and very effective. If an image has a color cast, you can use Photo Filter to create a counteracting color cast. Or if you feel an image would benefit from an overall hint of color (e.g., warmth), you can apply that as subtly or aggressively as you’d like.

In this example, I masked the sky to warm up only the buildings, and I disabled Preserve Luminosity, allowing the buildings to get darker and more richly colored.

I sometimes use this adjustment to figure out just what color permeates my image. If I suspect the color cast is greenish, I’ll try a Magenta filter and play with its Density. Double-clicking the word Density resets it back to its default value of 25%.

Selective Color

This is yet another adjustment that allows you to alter the color of individual hues. It is built to be friendly to those accustomed to thinking in CMYK, whether they’re in CMYK mode or not.

You start by choosing a color (including non-colors like white, black, and neutrals) from the Colors menu. Then you can use the four sliders to adjust that color. Each slider affects (or simulates affecting) a process color (Cyan, Magenta, Yellow, or blacK) by adding or removing ink from that plate. In RGB, it acts the same way without having to commit to a specific CMYK color space. For more on why that’s wise, see “Adjusting in RGB or CMYK or Lab” (page 218).

For example, if you want greens in an image to be brighter green, you can remove Magenta. The buttons at the bottom of the adjustment, Relative and Absolute, affect how the addition or subtraction of “ink” is calculated. Remember that process colors are controlled by percentages: all set to 0% is just white paper; 100% is the maximum for each. Absolute will alter the amount of that ink exactly as the slider indicates. So if an area used 50% magenta, an adjustment of -20% will change it to 30%. Relative is proportional to what’s already there. So that -20% will remove 20% of 50, or 10%, resulting in that area using 40% magenta.

Channel Mixer

Because people are often given the bad advice of converting to CMYK before they know which color space they really need, the resulting file often needs work. Although Channel Mixer can be used in RGB, it seldom is. Most often, it’s used in CMYK by prepress professionals to rehabilitate a document, giving one process color plate more substance by borrowing from another, for example:

Camera Raw Filter

Several “adjustments” can be applied only like filters, including this one. I strongly recommend converting the layer(s) to be adjusted to Smart Objects first to perform these adjustments nondestructively.

This filter can be applied using Filter > Camera Raw Filter…. It’s prodigious capabilities are covered in the chapter “7 Editing Raw Files” (page 350).

Shadows/Highlights

This adjustment attempts to recover nearly lost highlight and/or shadow detail with minimal effect on the other or on midtones. It is actually a filter in disguise, hidden in the Image > Adjustments menu. Thus, you should convert the layer(s) to be adjusted to a Smart Object to perform this adjustment nondestructively. When the Shadows/Highlights dialog is open, be sure to check Show More Options to access the adjustment’s full capability.

The dialog has three sections: Shadows, Highlights, and Adjustments. The first, Shadows, lightens tones just barely lighter than black, while Highlights darkens tones just barely darker than white. Within each of those first two are three sliders:

• Intuitively, the Amount sliders control how much each adjustment lightens or darkens.
• The Tone sliders control how far into the tonal range each adjustment extends. Thus, if Tone is set to 35%, the effect covers about a third of the full tonal range. I find that particular value a better starting point than the default of 50%.
• Finally, the tricky Radius sliders. These define the tonal “regions” affected. Large expanses of shadows, for example, may require a Radius in the hundreds of pixels. Thin, narrow areas may be happier with a setting in the tens of pixels. I like the advice of the wise Dr. Russell Brown (russellbrown.com): “Move the Radius slider up and down until the image looks better—then stop!” This simple advice works because when the radius value is too high or low, there is either too little effect to notice or obvious halos around objects.

Since shadows and highlights lack saturation, the Color slider in the Adjustments section adds some. Midtone adds contrast to the middle of the tonal range, which may have been lost in this adjustment. It may cause you to lose some of your gains, however. A small number of pixels should likely be allowed to clip, so the 0.01% values are good defaults.

HDR Toning

HDR Toning is not actually an adjustment. Like Shadows/Highlights, it is a filter. Be warned that it is extraordinarily destructive. Not only can it not be applied to a Smart Object, the document to which it’s applied must be flattened to a single Background layer!

It can be used to recover nearly lost shadow or highlight detail and give those areas pleasing saturation, but it can also make an image obnoxiously surreal.

Use this filter with discretion and only on a copy of your original files.

Replace Color

This adjustment cannot be applied as an adjustment layer nor as a Smart Filter. It may have some appeal, as it combines the interface of a Color Range selection to choose a color and the three sliders of a Hue/Saturation adjustment to change that color. But since it can be used only directly on pixels, I’d prefer to use those other tools individually and nondestructively.

Equalize

Recall that a histogram shows the relative number of pixels each level (shade of gray) has on each channel. Typically, some levels have fewer pixels to represent them, and others have more. Equalize tries to give each level, on each color channel, the same number of pixels, resulting in a rather level histogram. The darkest tone is made black and the lightest, white.

If there are already many black or white pixels, you may still see a spike at one end of the histogram or the other.

If the image started off rather light, the result will be darker, or if it started dark, the image will be lightened. To force a result, you may start with a selection. If there’s an active selection, a dialog box opens asking whether to apply Equalize to only that area, or to use the tones in that area as the basis of the adjustment that affects the whole image.

I rarely use this adjustment because its uses are limited and it cannot be applied as an adjustment layer or Smart Filter. I sometimes find it useful on a copy of a document before using the Threshold adjustment (see page 248).

Creative Adjustments

Again, I’m listing the following adjustments here because they’re most often used for more creative, interpretive results. Of course, if that’s what you do day-to-day, then they’re wonderfully practical production tools too.

Black & White

The name nearly says it all. With this adjustment, especially as an adjustment layer, we seemingly discard color but translate the image’s existing hues into lighter or darker tones as we choose.

Invert

This adjustment is very simple: it creates a “negative” version of the image. All colors and tones become their opposites. Using a blend mode like Color or Luminosity, you can restrict the effect to those attributes.

Posterize

Posterize reduces the number of levels or tones in an image. If the image is grayscale, the number set for Levels is exactly how many shades will be left. In color images, each channel is a grayscale image and is reduced to the specified number of levels. In either case, gradations become visibly “banded.”

Often, using another adjustment first (or an adjustment layer below) can yield a more pleasing or seemingly detailed result. I often use the Equalize adjustment on a copy of the layer my image occupies before creating a Posterize adjustment layer.

Threshold

The result of Threshold is initially very similar to applying a Posterize adjustment to a grayscale version of your image with Levels set to 2. That is, you’re left with two shades of gray: black and white—nothing else. This is often done to create a stenciled look. Unlike Posterize, Threshold allows you to choose the Level below which becomes black and above which becomes white.

Gradient Map

This is a powerful adjustment for “toning” images in way that is reminiscent of darkroom processes. Of course, as this is Photoshop, we can go much further. With this adjustment we apply a gradient whose colors “map” to levels (shades of gray) in the image. The color at the left edge of the gradient preview replaces black in the image, the color at the right edge replaces white, and, as you’d guess, any color in between those replaces tones in between. Look at the following silly example:

In this case, a purple has been mapped to black, orange is mapped to tones a little lighter than midtones, and cyan is mapped to white. Below are more plausible examples. One emulates a cold-toned image on warm (yellowish) paper and the other a more typical “sepia” look.

You should review how to control the Gradient Editor to better utilize this adjustment. See “Gradient” (page 137).

Color Lookup

Color Lookup is another “mapping” adjustment. Unlike Gradient Map, which maps colors only to tones, Color Lookup can map colors to colors, to quickly achieve a consistent look and feel, even one that is complex.

This type of adjustment was created by the film industry, which has to deal with imagery from many sources in many different formats. By applying a color lookup table, or “LUT,” they can match raw footage from various sources to acheive the desired look.

However, the main purpose of LUTs is not color correction, but creative color alteration. In the world of cinema, this is called “grading,” and using LUTs is part of it.

Using a LUT

To achieve reliable results, an image should start well exposed (neither too light nor too dark) and properly white balanced. See “White Balance” (page 354) for more on that. Then a LUT can help us achieve consistent appearances, whether that’s a film-stock emulation, a special effect (like night from day), or a gentle color/tonal shift. Note that I said it “can.”

LUTs come in several basic flavors: 3DLUT, where the “3D” refers to the three color channels (RGB); DeviceLink, which are RGB or CMYK color profiles, the latter having four dimensions but still confusingly called 3DLUTs; and, finally, Abstract profiles, the only ones that are color calibrated, but which few software applications use. The first two kinds will look different in different color spaces so will generally need a tweak to match when images come from different sources. The last is wonderfully consistent since it adapts to the current color space to match actual color.

So, if you’re working exclusively within Photoshop, try to use Abstract LUTs. If you use other applications, too, then you may have to use 3DLUTs since they’re more widely used. Because of their near ubiquity, you’ll likely find more 3DLUTs to use.

Choose the Color Lookup adjustment from the Adjustments panel, then choose one of the many LUTs that come with Photoshop. You may use one you’ve made or downloaded, too, by choosing to Load it from the appropriate menu (.cube, .look, and .3dl are 3DLUTs, for example, but DeviceLink and Abstract are ICC profiles with .icc extensions and may be hard to distinguish from one another).

Creating a LUT

Creating your own color lookup table (LUT) is not hard. It can benefit from foresight, however. No matter which kind of LUT you’re creating, you need to start with a simple Background layer and no others. Then use adjustment layers to create the look you want to achieve with the LUT. Although you may use color fill layers with blend modes as well, that is less reliable.

If your objective is to create LUTs for use in Photoshop alone, you should create Abstract profile LUTs to ensure optimal color accuracy and consistency. To do so, the document needs to start not only with a Background layer, but it needs to be in Lab Color mode (Image > Mode > Lab Color). In fact, this is the only reason most of us will ever have to use Lab mode! Not all adjustment layers are available in Lab, and those that are will behave slightly differently than they do in RGB mode. Abstract profile LUTs can be used in all color modes too!

If you’re in prepress and are compelled to work in CMYK, you’ll need the exported DeviceLink profile (the only one exported that doesn’t have “RGB” appended to its name). To use a LUT in video applications, it’s most likely you’ll need a 3DLUT.

• Consider using a duplicate of your original image.
• Start with a Background layer.
• For color-accurate Abstract profile LUTs, convert to Lab Color mode. The exported LUT without “RGB” appended to its name is the Abstract profile. Otherwise, working in RGB is fine.
• Create adjustment layers with intact (but unused) masks.
• Choose File > Export > Color Lookup Tables….
• Provide a Description and, if desired, copyright attribution (“© Copyright” and the year are automatically prepended to this).
• Choose the Quality (accuracy) and Formats you want. For the former, 64 is considered a good accuracy/file size compromise, and I check all of the latter and weed them out later.
• Click OK to choose a location and base name for the exported LUT files.