“At B&H’s Optic 2018, photographer Seth Resnick discusses color, enhancing creativity, and processes you can employ to make better images and improve your photography. He stresses the importance of being aware of the details in your environment and mentally taking note of what you see BEFORE you start shooting with your camera.”
Learn more about Seth Resnick here.
Find out more about our Digital Photo Destinations Workshops here.
When you want to make a selection based on hue, start with Photoshop’s Color Range. (It’s located in the Select menu.) It’s an invaluable selection tool that’s easy to master.
The heart of this tool is the default setting for the Select drop-down menu – Sampled Color. Once you learn to use it, you’ll find you’ll use it often. For Sampled Colors it’s an oversight to activate the slider for Fuzziness (the number of related hues included) and not Range (targeting specific lightnesses). Hope, no request, that Adobe activates both Fuzziness and Range sliders for all drop down settings. For now, you can overcome this limitation to some extent and customize any range of color with surprising precision by using the icons on the right of the dialog box Eyedropper Tool, Add to Sample, and Subtract from Sample icons as well as the Invert checkbox. It is also the only setting that activates the Localized Color Clusters check box, which essentially adds a radial gradient around the point you sample a color from. You can master this tool in a few minutes.
You’ll probably find that you’ll use the other settings in Color Range’s Select drop-down menu sparingly, some not at all.
Skip them. The color choices in the Color Range drop-down menu – Reds, Yellows, Greens, Cyans, Blues, Magentas – are almost useless. They’re not as accurate as you’d like them to be and they don’t offer Fuzziness or Range sliders to control them with. It’s all or nothing, usually nothing.
Use them occasionally. The Highlights, Midtones, and Shadows options have improved by activating both Fuzziness and Range sliders. Curiously, if you Invert the Highlights setting you get a slightly different result than simply choosing Shadows setting and vice versa. But don’t worry, the Fuzziness and Range sliders will give you all the control you need to compensate and much, much more. Color Range’s selections of luminance yield different results than making luminance selections by loading channels as selections (See my article Masking Luminosity.); for better or for worse, they tend to produce fewer gray values and so they yield more generous selections with quicker less smooth transitions into surrounding values, as if you added contrast to a channel selection.
Consider it. While it offers only the control of the Fuzziness slider and not the Range slider but adds a Detect Faces feature, Skin Tones can be quiet useful – at times. It is clearly biased towards Caucasian skin tones as it picks up whites before darker browns but it does a good job of avoiding very saturated warm hues. If Skin Tones fails, use the default Sampled Colors instead and choose a custom base color you’d like to start with.
Forget about it. Be careful about the Out Of Gamut feature. It works based on the profile loaded for an output device, usually an offset press. It’s designed to help you prepare files for printing by selecting and subsequently desaturating colors that are too saturated to be printed accurately. Using color management and good output profiles is a better way to control gamut compression.
Finally, if you want a larger preview of the selection/mask being generated, the Selection Preview drop-down menu offers four settings that will change the appearance of the image window Grayscale, Black Matte, White Matte, and Quick Mask. In most cases, the generously sized icon in the Color Range window will be all you need.
Photoshop’s Color Range is an indispensable selection tool that continues to improve. When you want to make a selection based on hue, start here.
Read more about masking here.
View more in my DVD Drawing With Light.
Learn more in my digital photography and digital printing workshops.
Dehaze may create color artifacts
Color artifacts removed
Color without Dehaze blended with luminosity with Dehaze
The top layer is set to a blend mode of Color
When you’re using Lightroom or Camera Raw, you’ll quickly find the Dehaze slider can produce marvelous contrast effects. Dehaze can dramatically exceed the contrast that can be produced with either Curves or Clarity. Sometimes it will reveal detail you couldn’t see with your eyes!
Often, there’s a price to pay for these great effects – color shifts. Neutral areas may turn magenta. Shadows may pick up strong blue or green casts. To make matters worse, these unwanted artifacts are rarely uniform, which makes them harder to fix.
If you’re lucky you can compensate by reducing Saturation after using Dehaze. When you do this, it’s likely that you’ll end up choosing the least objectionable version or making a compromise you’d prefer not to. Frequently, to avoid these side effects, you’ll be tempted to not to push Dehaze as far as you’d like to.
There is a cure that will help you go as far as you’d like, without producing color shifts. Render your image twice. First, render it with as much Dehaze as you’d like. Second, render it without Dehaze.
Then place the version without Dehaze in a layer on top of the version with Dehaze. Change the Blend Mode of the top layer to Color. This will give you a combination of the color of the top layer (without Dehaze’s color artifacts) and the luminosity of the bottom layer (with Dehaze’s contrast).
How do you make two layers from one Raw file?
If you’re using Lightroom, make a virtual copy and then double click on the Dehaze slider. Highlight the original file and the virtual copy and select Photo > Edit In > Open as Layers in Photoshop. Now in Photoshop, make sure to change the top layer’s blend mode from Normal to Color.
If you’re using Camera Raw, open your Raw file as a smart object, then select New Smart Object via Copy in the Layer menu, and finally double click on the top layer to return the Dehaze slider to 0. Remember, change the top layer’s blend mode from Normal to Color.
The technique of using the color of one layer to overlay another layer can be used for many applications. Here, it makes Dehaze even more useful.
Check your inboxes!
My enews Insights is out.
This issue features many valuable resources on Color Theory.
Sign up here.
Remember, your username is your email. Your password is free.
What is color theory? It’s a language and set of concepts that’s useful for describing and creating color relationships.
Color management helps you get consistent colors. Color adjustment helps you change colors. Color theory helps you choose colors.
I’ve studied color all of my life – and it’s always rewarding.
Here’s a roundup of the best books I’ve found on color theory.
Lori Reid’s The Color Box is a great place to start with color theory. It’s neither definitive nor deep but it does a great job of surveying many approaches to understanding color.
Josef Albers’s Theory and Interaction of Color is the definitive work for creating a hands-on understanding of essential optical effects.
Johannes Itten’s Art Of Color (an expansion of his The Elements Of Color) is clearer and wider-ranging than Albers’ similar book.
Faber Birren’s Creative Color discusses ways of creating effects (like iridescence and pearlescence) that few other resources touch on.
Wucius Wong’s Principles Of Color Design shows the links between color and other compositional devices.
Margaret Livingstone’s Vision And Art details the biology behind the effects artists create.
An essential quality of color is temperature. Temperature can be used to attain a color balance. Temperature can be used to enhance spatial relationships within an image. Temperature can be used to elicit psychological responses within the viewer. Understanding and exploring the dynamics of temperature in color can benefit any visual artist.
There are physical characteristics of color linked to temperature. The color temperature of light (Kelvin degrees) is determined by measuring a black body radiator (an object heated so that it emits light). As the physical temperature of the object rises, color transitions from red (long wavelengths – low energy) to blue (short wavelengths – high energy) through ROYGBIV (red, orange, yellow, green, blue, indigo, violet). When it comes to light sources, physically, blue is warmer than red.
There are also psychological qualities of color linked to temperature. Psychologically, blue is cooler than red. These associative qualities of color with regard to temperature are almost universally accepted. This is due in large part to our physical environment – water is blue, plants are green, sunshine is yellow, fire is red.
Using the qualities of one sense (touch) to describe the qualities of another (sight) can be a tenuous affair and may lead to ambiguity and confusion. The more precise a language is the more useful it is. The language of HSL (hue, saturation, luminosity) is a very precise language. When using the language of HSL, hue values mark a position measured in degrees on a color wheel. A circle has 360 degrees, so the scale is 0 – 359.
While every degree represents a new hue, you can use broader terms to describe a color family; red, orange, yellow, etc. Think of the color wheel as a clock where every hour marks a new color family.
90 yellow green
150 blue green
210 green blue
330 blue red
Absolutely warm and cool colors can be found at 0 (red – the warmest color) and 180 (cyan – the coolest color) degrees. Determining whether one color is warmer or cooler than another can be measured by their proximities to these poles. A line between 90 (green yellow) and 270 degrees can be used to broadly demarcate warm colors from cool colors; colors on the right (towards red) are warm while colors on the left (towards cyan) are cool. The association of yellow with the sun, a warm light source, subtly skews the associative quality of warmth towards yellow (60) and away from blue; as a result, colors above the line between 0 and 180 tend to seem warmer than colors below it. (i.e. while both are equally distant from red (0), orange (30) seems warmer than blue red (330).) While one color can be seen as warmer or cooler than another color, each color also has warm and cool components; there are warm yellows and cool yellows, warm blues and cool blues, etc. (Where numerical classifications of colors define hues very specifically (1 degree per hue, 30 degree spread per linguistic color), linguistic specifications of colors (red, orange, yellow, etc) define broad ranges of hues.) Defining the warm and cool endpoints of any linguistic color is useful at a coarse level of granularity but becomes increasingly subjective at a fine degree of granularity. At what point does blue become purple? At what point does blue become green?
It’s possible to describe the adjustment of hue simply in terms of warming and cooling.
Photographic color adjustment strategies rely on adjusting a balance in each of three complements.
Red – Cyan
Green – Magenta
Blue – Yellow
Most hue adjustment tools, like Photoshop’s Color Balance, have these complements built into their interface. You can’t increase one hue without decreasing its complement.
Each set of complements has a warm and cool dynamic.
R (warm) – C (cool)
G (cool) – M (warm)
B (cool) – Y (warm)
These three complementary axes have different warm/cool dynamics with respect to the three color primaries – RGB.
red (warm red) – cyan (cool blue) warm/cool
green (cool green) – magenta (cool red) cool/cool
blue (warm blue) – yellow (warm green) warm/warm
You can analyze the color temperature dynamics at work in any image by sampling it and graphing it. Doing this will not only help you understand how it works but also how you might improve it.
You can make a field of color appear more dynamic, complex, and three-dimensional by preserving or introducing a variety of warm and cool components in it.
The temperature of color carries spatial associations with it. Warm colors tend to appear to be nearer than cool colors. Again this is universal. It can be overturned by many factors; some factors are related to color, such saturated colors appear nearer than desaturated colors or a progression from light to dark may be the primary element that establishes spatial hierarchy; some factors are not related to color, for instance, placement and overlap in composition may be primary spatially, overriding color relationships.
Color balance, spatial proximity, association – these are just three of the uses of warm and cool color dynamics in images. Whether you are adjusting preexisting color relationships or creating new ones, having thoroughly explored the warm and cool dynamics of hue, you can apply that knowledge towards the realization and enhancement of your images.
Learn more in my digital printing and digital photography workshops.
Learn more with my DVDs on Color here.
Learn more with my free color resources here.
Graphing color can help you identify colors more specifically and understand relationships between color better. One way to graph color is to use the Color Sampler Tool in Photoshop in combination with the Apple color picker.
Using The Apple Color Picker
To access the Apple color picker choose Photoshop: General Preferences and change Color Picker from Adobe to Apple and click OK. To see the new the color picker click on Set Foreground Color or Set Background Color icon in the Tool bar.
Unlike the Adobe color picker, the Apple color picker is a color wheel. Creating and using color wheels to describe color and plot color relationships is a time honored tradition dating back to Leonardo DaVinci. Some of the most famous color wheels were created by Newton, Goethe, and Munsell. The Apple color picker is an additive color wheel where complements are defined as red and cyan, green and magenta, and blue and yellow.
You can sample any color in an image and find its position on the Apple color wheel. Using the Eyedropper Tool, sample a color in a composition. Then click on the Set Foreground Color icon. The Apple color wheel will appear and a small circle will plot the sampled color. You can make a record of this chart by taking a screenshot of the color wheel (caps lock, Shift, Command, 4). This will create a document on your desktop, which can be opened in Photoshop.
Combining Multiple Samples
You can combine multiple sample points into a single chart by taking multiple screenshots, opening them in Photoshop, and combining them. Drag the Background layer from one document into another and give it a meaningful title. Make sure the two layers are registered with one another. Then, mask off everything on the top layer except the circle identifying the color on the color wheel, the triangle identifying its luminosity on the slider to the right of the color wheel, and a portion of the color bar above the color wheel. You’ve just graphed the two colors on the Apple color wheel. You can do this with as many colors as you desire.
Once colors have been graphed you will be able to identify a variety of relationships between colors, both colors that exist in a composition and colors that do not.
Sidebar For Luminosity
You can identify luminosity values with the slider to the right of the wheel. By sampling colors from an image and noting their position on the slider you can determine whether an image is light (high key), medium, or dark (low key) and whether it contains low, medium, or high contrast, by comparing the distance between sampled points (the greater the distance the greater the contrast).
Lines For Differences In Hue
You can draw a straight and/or a curved line between two hues. This will help identify all the other hues between the two, useful in creating gradations or smoother transitions between two colors.
Concentric Circles For Saturation
You can draw concentric circles with varying radii on the wheel. This will help define the saturation level of an image – neutral, semi-neutral, saturated, or supersaturated.
Triangles For Color Families
You can divide the color wheel into triangular slices of varying widths. This will help define hue relationships within an image – monochromatic (single family hues)(within 30 degrees), analogous (closely related hues)(within 60 degrees), complementary (opposite hues)(separated by 180 degrees), or split complementary colors (three colors where two are found between 150 – 210 degrees from the third).
Shapes Show Relationships
You can draw geometric shapes inside the color wheel – triangles, squares, rectangles, pentagons, hexagons, octagons. This will help identify hues that may be used to create a logical color structure – diads, triads, quadrads, etc.
A little exploration and mapping of color will help you make many new discoveries about the color dynamics in any image, both of how they work and how they can be improved.
Here are a few observations about the color dynamic revealed by graphing it.
1 The composition has four significant colors.
2 The luminosity has only moderate contrast, constrained to the mid-tones, weighted low.
3 The palette is reasonably saturated, but not supersaturated. The most saturated hue is dark purple, the second most saturated color is pink.
4 A 60 degree slice reveals that there are two reds and two purples – analogous (closely related) hues. The brown is actually a dark red with relatively low saturation. The reds are closely related in hue; so are the purples. There is a light set and a dark set of reds and purples. The hues are at the boundaries of analogous relationships, providing maximum variety within that color dynamic. Contrast is derived based on luminosity first, hue second, and saturation third.
5 While one hue, red, is the warmest color, the other hue, purple, lies close to the boundary between warm and cool colors. There isn’t a strong contrast between warm and cool, but there is a great deal of variety within a predominantly warm composition.
6 Gradation (smooth transitions) can be accentuated by including hues in between the sampled points either along a straight or curved line. The background color for the sky seems continuous and smooth because it contains many small steps between the lighter hue at the top and the darker hue at the bottom; this has a calm effect. The pink accents in the highlights appear warmer and lighter because they are surrounded by cool dark colors, without substantial gradation between them; this has an energetic effect.
One of the most distinctive features of a visual artist’s use of color is their use of saturation.
Many photographers are often asked, “Are you a black and white or color photographer?” (Curiously other visual artist’s are rarely asked this question.) While many people who ask it don’t mean it to be, it’s a loaded question. There’s often a latent assumption that you can’t do both well. In fact, work with one can strengthen work with another. Moreover, the question suggests that black and white (and shades of gray) are not colors, when in fact they are very specific colors – neutral colors. And, the question does not address with any specificity how a photographer uses more saturated color. Curiously, this question is rarely asked of painters and filmmakers. A more useful question might be, “How saturated is your palette?”
There are essentially six distinct levels of saturation – neutral, semi-neutral, reduced saturation, fully saturated, highly saturated, and super-saturated.
Neutral colors are characterized by a lack of saturation. Neutral lie along the luminosity spine that unites all hues. (Equal parts of complementary colors mix to create neutral colors.) At any one luminosity value there is only one truly neutral color. Using neutrals is comparatively simple, as two other variables (hue and saturation) in color are reduced to the same values (zero); it’s an economy of means. To appear luminous neutral images typically employ significant contrast in luminosity to compensate for it’s lack of hue and saturation. With a few exceptions, neutral images generally present a radical transformation of the way we see. Many people associate it with dream states, whether they dream in black and white or not. It is often thought of as elegant, if not conservative and restrained. The luminosity component of color aids in judging levels of illumination, spatial relationships, and volume. Formal qualities and qualities of illumination become a particularly significant concern in neutral images. Neutral images are often associated with antiquity or of timelessness.
A semi-neutral colors contain trace amounts of saturation. Values cluster around the luminosity spine within a small radius. Semi-neutrals do not include truly neutral colors. Semi-neutrals function very similarly to neutrals with additional complexity. Spatial relationships may be subtly accentuated by the slight addition of hue. Iridescent and lustrous color effects can also be achieved.
Images with reduced saturation use low levels of saturation. Values cluster around the luminosity spine within a larger radius but do not achieve full naturalistic saturation. Though not fully presented, the hue of elements within a composition becomes apparent. Similarly the ambient color temperature can be suggested. Spatial relationships can be more clearly delineated through hue. Hue becomes a significant concern in compositions, though it may or may not be a primary concern, depending on how it’s deliberate reduction is handled. Images with reduced saturation often do not appear realistic. They appear somewhat restrained, limnal, neither here nor there, but somewhere in between. Many people associate reduced saturation with the past, particularly the recent past.
Fully saturated images are the most representational images. Color can be used to make elements in a composition seem similar or different depending on the degree of contrast of one or more of its elements – hue, saturation, or brightness. All three elements of color can be used to describe qualities of form, space, and light. Often, to preserve a realistic appearance, luminosity contrast is lower than in neutral and semi-neutral palettes. Fully saturated palettes are relatively energetic. They have decidedly contemporary connotations.
Highly saturated images use a generous amount of saturation, without becoming unrealistically saturated. Hue becomes a dominant preoccupation. It may vie for becoming the primary focus of attention with subject, spatial relationships, or qualities of light. Highly saturated images are energetic. They have contemporary, generally expressive, and sometimes youthful connotations.
Super-saturated images use extreme levels of saturation. Realistic representation is generally not a concern. Colors often appear unrealistic. Additionally there is a tendency for colors to posterize, with abrupt transitions between different hues. Hue becomes the dominant concern, over subject or spatial relationships. Super-saturated images tend towards flatness. They are highly energetic, expressive, and sometimes associated with altered states of consciousness.
You can use the Color Sampler Tool in Photoshop in combination with the Apple color picker to graph color and identify the level of saturation of one or more images, either yours or someone else’s. See (list url for graphing color article) to find out how.
To lend unity to images within a single body of work it can be helpful, and in some cases necessary, to limit your use of saturation. Many artists will use just one level of saturation. While it is extremely difficult to present a wide array of saturation levels within a single body of work (though not a single image), for instance both neutral and saturated images, it is possible to present images that employ more than one level of saturation if they are closely related, such as reduced saturation and fully saturated or fully saturated and highly saturated palettes. As variety in saturation level rises it becomes more difficult to have separate images be seen as related.
When you think of Ansel Adams’ photographs you think of neutral images rather than highly saturated ones. When you think of Matisse’s paintings you think of supersaturated images rather than neutral ones. Think of your use of saturation as an essential component to defining your own signature style.
Learn more in my digital printing and digital photography workshops.
Learn more with my DVDs on Color here.
Learn more with my free color resources here.
In this video, I share some of my feelings about photography and thoughts on color.
Find out more about X-Rite here.
Read more about color management here.
Learn more in my photography and printing workshops.
Get my free Gear Guide here.
Before channel blending
After channel blending
Big problems call for big solutions. Blending channels is a powerful color adjustment strategy that can handle even the biggest challenges.
Color adjustment occurs by modifying the tonal structure of individual grayscale channels. Typically, the information within them is adjusted. Less typically, the information within them is replaced.
Blending channels is one way of replacing them. Blending channels takes information from one channel and combines it with information from another. Rather than simply enhancing existing tonal values, blending channels reshapes one channel’s tonal structure with another’s. Consequently, in a most cases, blending channels calls for a substitution of information by percentage not a wholesale replacement of the deficient channel. You usually blend channels from different versions of the same image because blending channels from different compositions produces a highly altered effect.
Blending channels is complex. It often produces additional unintended color affects that may require further correction, such as shifts in hue that aren’t uniform across the tonal scale. Blending channels is neither the simplest nor the most direct path to color adjustment, but in certain situations (files that are exceptionally problematic) it may be the best path. The resulting benefits can be dramatic.
When is blending channels appropriate? In extreme cases. Blending channels is designed to correct major color deficiencies. It’s recommended if a channel is severely deficient, either globally or in select areas. For example, by being extremely light or dark or having very low contrast, a channel may be lacking desired detail. That detail can be found in another channel. Fine-tuning color is best left to more traditional methods of color adjustment.
There are several ways to blend channels; Channel Mixer, Apply Image, Calculations, and using channels as layers. Let’s review the options in detail.
The Channel Mixer (Layer: New Adjustment Layer: Channel Mixer) blends percentages of channels within a single document. It can be applied as an adjustment layer and so corrections made this way can be changed or masked indefinitely. It cannot be used to blend channels from two documents. The Channel Mixer is an excellent choice for making global (the same percentage of channels for the whole image) color to black-and-white conversions. If you want to control black-and-white conversions locally (different percentages of channels for different image areas), use channels as layers instead.
The commands Calculations (Image: Calculations) and Apply Image (Image: Apply Image) can also be used to blend channels. With these two commands you can combine any two channels, from different documents, from any layer, at any opacity, with most blend modes. With Apply Image you target the channel you wish to change. With Calculations you blend to create a new document, a new channel, or a new selection. Neither Calculations nor Apply Image can be used as adjustment layers or layers, consequently corrections you make with either of these features are made permanently to an image. With Apply Image and Calculations you can take advantage of two less frequently used blending modes not found with other tools (Add and Subtract) but you cannot take advantage of four frequently used blending modes (Hue, Saturation, Color, and Luminosity) – even if you use the Fade command.
For the greatest control and flexibility use channels as layers. How do you do this? Copy any channel and paste it into any destination as a layer. (Target a channel (click on it); copy it (Select: All; Edit: Copy); then target the master channel (RGB) and paste (Edit: Paste).) You can activate, deactivate, mask, change, or replace this new layer indefinitely. Use Layer Styles (double click on the layer icon in the Layers palette) to determine Blend Mode, Opacity, Advanced Blending (to select which channel is affected) and Blend If options (to determine how This Layer affects the Underlying Layer or which values of the overlying layer affect the values of the underlying layer). What’s more, you get a dynamic preview of any changes you make while you make them. The adjustments you make are flexible, so you can remove them or fine-tune any of the settings future editing sessions. You can even blend two or more channels first, as layers, and then use the resulting new layer to blend with the Background layer. By turning channels into layers, you can achieve everything that the other methods achieve and more.
One File, Many Channels
You may be surprised to find that every file has at least ten channels to choose from. How do you get so many? Consider the file in different color spaces – RGB, CMYK, and LAB. Convert a duplicate file into another color space and you can use any and all of the resulting channels. In fact, you can choose between many, many more channels when you consider that when converting to CMYK there are five different options for generating a Black Plate (None, Light, Medium, Heavy, and Maximum) with two styles for each with two Separation Types (UCR and GCR). But, for the vast majority of situations, I recommend you try to keep things as simple as possible and stick with the standard three.
Be cautious with older files and lower end scanners when blending with the blue channel as it often contains significant amounts of noise. In fact, in some instances, blending channels can be used to replace some or all of the blue channel and thereby remove unwanted noise. Unlike blurring or despeckling, this method of removing noise will not compromise sharpness, but it may produce unwanted color shifts that will require subsequent correction.
A Good Preview
The possibilities are staggering. Is there anything that can help with the decision making process? Yes. A good preview. You’ll want to have multiple documents of the same image in different color modes (RGB, CMYK, LAB) visible at one time to simultaneously see the blended and the blendee. You may even want to make a side-by-side comparison of the component channels of a single document. To do this, use the Split Channels option in the Channels palette submenu. This command will break a single multi-channel document into multiple single-channel documents. (If a file has layers it must be flattened first to use Split Channels.) While doing this with several documents will quickly fill a screen, having the channels separated makes evaluating their relative merits infinitely easier.
With so many possibilities, how do you choose one channel as the best candidate to blend with another and how do you use it?
First, identify the channel causing the problem. Then, find the channel with the best contrast in the areas you wish to enhance (at a low opacity) or replace (at 100% opacity). (Stronger adjustments require higher opacities.) Finally, deal with any unintended side effects.
There are several tried and true strategies for dealing with classic problems. First, create detail where there was none before. Second, create contrast that wasn’t there before. Third, add more contrast to existing tonal relationships, if the values can’t be adequately enhanced using Curves.
Look to the Luminosity channel in LAB. Look to the black plate in CMYK. Look to complementary colors. Complementary colors often contain the best possibilities for increasing contrast (Red and Cyan, Green and Magenta, Yellow and Blue), in highly saturated values.
What are you looking for? Better detail in shadows or highlights, better contrast, and a similar tonal distribution. If you change the relative distribution of tones in a channel, you’ll create a non-uniform color shift where some colors will shift more dramatically than others.
Sometimes channel blending produces unintended side effects.
In most cases, these side effects can be cured.
Blending channels can produce unintended side effects. There are times when it’s better to achieve the necessary effect with this technique and accept its side effects, if the side effects are easier to cure than the initial problem. Typically, all that’s required is a little dose of additional tonal enhancement, either to the master channel (tone and contrast) or a single channel (color). If you find this is not the case, take this as a sign that this is not the right technique for the problem you face.
Layers offer many Blend Modes
Control The Mix With Blend Modes
As well as controlling the amount channels are blended you can control the way they are blended, by using blending modes. Blend modes determine how new values are mixed with old values. There are dozens of blend modes to choose from.
As color adjustment is achieved by altering the luminance (light and dark) values of select channels (Channels create but don’t contain color or saturation.), when it comes to blending channels, you can limit the number of blend modes you use to those that affect tone; five are particularly useful – Lighten, Screen, Darken, Multiply, and Luminosity.
Lighten displays the lightest values of both This Layer and the Underlying Layer; its neutral color is black (you can’t lighten with black).
Screen multiplies the inverse values of the pixels lightness or darkness. It’s like registering same image in the same location from two projectors. Think of it as industrial strength lightening. Its neutral color is black (you can’t lighten with black). Screen can do wonders for opening up deep shadows. It has a tendency to blow out highlights. Use a contrast mask to remove the effect from highlights.
Darken displays the darkest values of both This Layer and the Underlying Layer; its neutral color is white (you can’t darken with white).
Multiply multiplies the values of the pixels on both layers and then divides by 255. It’s like registering two identical transparencies on a light table. Think of it as industrial strength darkening. Its neutral color is white (you can’t darken with white). Multiply can do wonders for reclaiming subtle highlight detail. It has a tendency to block up shadows. Use a contrast mask to remove the effect from shadows.
Luminosity combines the luminance values of This Layer with the hue and saturation of the Underlying Layer; it has no neutral color.
Enhance The Blend
You can enhance a channel before (or if you use channels as layers after) blending it with another. Use any adjustment method that makes the data better to blend with. As you’re blending with black-and-white images, Curves is usually all you need for it offers the most precise control of tone. For instance, you might increase the contrast of an image before using it to blend with. If you’re using the channels as layers method, all you have to do is group a Curves adjustment layer to the new layer being used to blend with. The contrast of overlying layer can then be fine-tuned as the blend with the underlying layer is occurring. This way you don’t have to guess how much contrast needs to be added before blending, instead you see how much contrast to add while the blend is occurring.
Constraining The Effect
While blending channels may solve problems that other adjustment methods can’t, they may also produce new problems.
In a great many cases, if the tonal distribution of a single channel is substantially altered using another channel, color may shift in an unintended manner. If this happens simply make an additional adjustment to eliminate any side effects. There are times when the color shifts you encounter will be non-uniform (more in some areas than others), which may lead you to making more complex corrections than you had anticipated.
If the problem solved with channel blending and the resulting side effects lie in different areas of the image, consider masking away the side effects rather than correcting them. There are several ways of masking the side effects of channel blending from selected areas. One, simply brush them away by painting with a black brush on a layer mask. Two, use a contrast mask to hold back the effect from highlights or shadows. Three, use the Blend If function in Layer Styles; by sliding the black arrow to the right or the white arrow to the left you drop out the effects from values below or above them – by holding the Option key (Command on PC) you can split the sliders to fade the effect more smoothly.
If you think you’re not used to blending channels, think again. Every time you turn a color image into a black-and-white image you blend channels. In a grayscale conversion three channels are blended to create a single channel while when using either Hue/Saturation or Channel Mixer three channels are blended to equal RGB values. But, when it comes to color adjustment, blending channels is used infrequently, perhaps because it’s so little known. Blending channels is a sophisticated adjustment method. In a majority of cases you don’t need a method that’s this complex. Blending channels is best used in exceptional situations for enhancing originals with substantial problems. If you find that you use this technique frequently, you’re probably not addressing the real problem, the quality of your originals. Nevertheless, when you run into files with severe problems, blending channels will often save the day.
Read more on Color Adjustment here.
Learn more in my digital printing and digital photography workshops.