How Color Works

Of course, that could be the title of an entire book or series. I’ll attempt to restrict myself to what you need to work in Photoshop. Please keep one thing in mind throughout this discussion: a color is what it looks like and not a bunch of numbers! That is, a color “happens” in your brain via the eyes. The trick is supplying your eyes and brain with the stimuli they need to recognize the color you want them to.

Color Modes

There are different ways in which we describe a color in Photoshop, and so Photoshop can operate in several color modes: RGB, CMYK, Lab, Grayscale, and a couple derivatives of these. Each mode attempts to work with color in the way different media or devices do. Grayscale, for example, is useful for images that will be output to devices that use only black ink. Although it may seem intuitive to work in the color mode of an anticipated output device, it is often far more optimal to work in one mode, and then create a copy for output. Which color mode should you use? You should choose one that offers more choices for color accuracy at output and is relatively intuitive. Let’s look at the candidates.

RGB and Photo History

RGB is about light in the darkness. No light: no color. By controlling the ratio of Red, Green, and Blue, we can create any color we want! Where both red and green light overlap, we get yellow. Green and blue combine to make cyan light. Magenta is the result of red and blue. All three make white. This is why we call these the additive primaries.

Note which color is absent or lacking. For example, cyan is the absence of red. When an image looks too magenta, we add a bit of green.

RGB has a rich and storied history. The theory that these three colors of light could be combined to make any color goes back to the early nineteenth century. In 1855, the physicist James Clerk Maxwell suggested the three-color photographic method that is used in nearly all processes today. A photographer named Thomas Sutton made the first color photograph using Maxwell’s method on his behalf. Three black-and-white images were made of the same subject (a tartan ribbon). Each was made through a colored filter: red, green, and blue. The resulting negatives were made into positive transparencies (“slides”). Readers familiar with something called film may feel mild nostalgia at reading that sentence.

Consider the scene above. If we were to photograph it through a red filter, the sky would appear dark, since there is little red light in it to pass through the filter, and the hills would be relatively light since they’re quite red. A blue filter would achieve nearly the opposite.

If we were to make each of those three images into black-and-white slides, they could be projected onto a screen simultaneously from three different projectors. The slide made from the red-filtered image would have red light projected through it, the “green” slide would use green light, and the “blue” slide blue light.

When the three projected images are aligned, the result on the screen will be a reconstruction of the colors in the scene! And color images have been made in more or less this way for about 160 years. Digital camera sensor arrays typically have millions of microscopic, monochromatic sensors with equally small red, green, or blue filters in front of them. The data from each get combined to give us the color image we edit in Photoshop. Even in the software, we can see this history: images in Photoshop have three color channels. They are each black-and-white images through which you’re to imagine red, green, and blue light projected onto the screen of your retina.

It gets tricky when we try to express how much of each of those colors we want. With most software, we use numbers between 0 (no light) and 255 (the maximum) for each. The RGB value of 0 0 0 is black—all lights out. However, as mentioned earlier, values like maximum red (255 R) on one device may not be as intensely red as on another. That is, 255 R will be a different color on different devices; maybe like Santa’s suit on one and more neon on another. After all, any device can produce only so great a range of colors. We refer to that range as the device’s “gamut,” which is described by an important data file called its profile. We’ll discuss profiles shortly.

So, we say that RGB values are device dependent. To accurately refer to a specific green, we might have to say something like “30 R 190 G 25 B as viewed on an Eizo CX270 monitor.” As we will see, a color-managed workflow that uses profiles deals with that for us. That’s one reason we call it “color managed.”

CMYK and Inks

CMYK is about dots of Cyan, Magenta, Yellow, and blacK ink on a substrate-like paper. Each ink absorbs a different color of light. Cyan, for example, absorbs red light, subtracting it from what gets to our eyes. This is why we call these the subtractive primaries.

Without ink, there’s only the white of the paper. By controlling the ratio of those four inks, we can make many colors. We measure the amount of each from 0% (none) to 100% (the full amount). In principle, we could use only cyan, magenta, and yellow, because the combination of all three inks should absorb all light, and thus appear black. However, this may be too much ink for many papers, or even if that much ink could be used, it may appear as dark brown instead of black. This is why we need black pigment too. Thus, CMYK is known as the four-color process, or “process color” for short.

Some printers use toner and some use ink. Those from different manufacturers produce somewhat different color. Paper also affects color, of course. Thus, every CMYK device is different. So, as with RGB, CMYK devices also have profiles that describe the range of color they can produce. That means that the CMYK numbers (or build) for any perceived color will vary from paper to paper, and from printer to printer.

Note: No single CMYK build looks the same on all printers.

Just as with RGB, CMYK numbers are device dependent: that is, they are meaningless without reference to a specific device. But again, the use of profiles in a color-managed workflow helps.

Profiles describe a device’s color space. What’s that? Since you asked…

Color Spaces, Profiles, and Lab

Since these terms are used so often, and often incorrectly, let’s be sure we understand what’s meant by them.

A color space can be compared to a physical one. Any position in a room can be mapped with the familiar x, y, z coordinates. Colors are mapped in a grand space called Lab, a mathematical model of human vision. L stands for “luminance,” with the a and b coordinates together mapping hue and saturation. Colors on the L-axis are gray. Colors farther from the L-axis are more saturated than those closer to it.

Sitting in a physical space, we can imagine the limited volume described by the points we can reach with our arms. A device (e.g., a printer or monitor) can produce only a limited “volume” of color within Lab space. That volume of colors is the specific device’s color space. Each unique device has its own unique color space. Lab itself is our color space: the range we can perceive.

In review, a color mode is the general way color is approached: either by adding three colors of light (RGB) or by subtracting light as it reflects off four colors of ink (CMYK). When I hear or read an instruction like “use the CMYK color space,” I cringe. I’m the annoying person who then asks, “which one?” A color space describes the color capability of a specific real or virtual device.

A profile is a data file that describes a color space. It is the means by which our software gets to know the color capabilities of a device. When looking at a list of color spaces, we are actually perusing a list of their profiles. Thus, folks often use the words “profile” and “color space” interchangeably, which is understandable.

Setting Working Space Profiles

Photoshop should be considered a device—it’s the virtual device we work in! Since it’s software, we can decide how large a color space it uses (how broad a range of color we can access). This also means that we are deciding the colors our color numbers refer to. Using Edit > Color Settings…, we choose settings that include this Working Color Space. That decision depends on workflow. We can choose from a number of different spaces ranging from the relatively small sRGB that approximates the color space of typical computer monitors, to the very large ProPhoto RGB space that encompass all the colors cameras can capture. There are two in between those: Adobe RGB, which attempts to closely match the color spaces of high-end printing presses on good paper, and P3, a “wide-gamut” monitor profile based on one used in digital cinema.

Depending on my needs, I may work in any of these, occasionally converting from one to another, changing the color numbers of every pixel. Luckily, computers do this math well and quickly, and can figure out which numbers to use for each color as we switch from one device profile to another.

Go to Edit > Color Settings… to see and configure your settings. I configured mine in both Photoshop and InDesign to help me in the preparation of this book. I began by choosing a preset called “Europe General Purpose 3” because it supplies the appropriate working spaces: sRGB for my illustrations and screenshots, and FOGRA39 since it is the CMYK profile for the press where this book was printed. I made a small adjustment to the RGB policy by choosing “Convert to Working RGB.” Many of the images I’ve used in this book may have used a different profile than sRGB, and I wanted them to all be the same. I chose to convert them to sRGB because if you are reading this as an ebook, sRGB is what the code inside that device expects. For print, those images in sRGB were converted to a CMYK color space profile, namely the working space profile. If I were creating only a printed book, a compelling argument could be made to use Adobe RGB for the RGB working space. But for this project, which targets many media, including e-readers, I decided sRGB was adequate and flexible. Other workflows, however, may require other choices.

For most Photoshop users, I’d recommend choosing the preset called “North America Prepress 2.” Although it will frequently cause disconcerting dialog boxes to appear when opening images (note the checked boxes for Profile Mismatches), it can provide the flexibility and color fidelity that photographers, designers, and illustrators desire.

The Color Management Policies for this preset attempt to retain the image’s color space profile even if it differs from Adobe RGB, the working color space of this preset. However, those Ask When checkboxes mean that you’ll have the option to convert to Adobe RGB when a file you’re opening uses something else.

Embedded Profiles

How does an image file “know” what profile it uses? Most of the time, that profile is embedded in the image so Photoshop (and other software) can correctly interpret its color numbers.

Converting vs Assigning Profiles

Two monitors may be sent the same signal to generate an intense red, but one will be more red than the other. This is just like when I turn the volume of an audio device up to 11: two different loudspeakers may produce different volumes of sound from that same signal.

To get identical results, I’d have to use a different signal for each of those different devices. That is, I may have to lower the volume to 9 for the louder speaker to match the other, or send different color numbers to the more vibrant monitor to match a less vibrant one. This is conversion. When we use Photoshop to convert an image to a printer’s profile, for example, it changes the color numbers of every pixel to the numbers required by that printer to achieve the same colors.

When we open a document that uses a color space profile that differs from our working space, and our Color Management Policies have been set to “Ask When Opening,” we can choose to convert it to our working space profile. Converting from one space to another changes the numbers to preserve the visual appearance of all the colors in an image.

In the case of an image downloaded from a website, it will likely be in the sRGB color space, whether the image editor who made it embedded that profile in the file or not. If there is an embedded profile, we’d see a dialog like the one above. If the embedded profile differs from our working space, most likely, we should convert it. In this way, if we sample a color from the image, we can use the same color numbers in InDesign or Illustrator (provided they, too, use the same working space), and we will achieve the same visual result.

If we’re editing an image provided by a colleague, we may choose to “Use the embedded profile” so we can return it in the same color space as it had when sent to us.

If we open an image with no embedded profile, we will have to guess which to use. Below, we see the dialog we’d get in that case. Almost always, the image is in the sRGB color space, but the profile wasn’t embedded for some reason. So, we choose to assign that profile to inform Photoshop what color each pixel is trying to be. We can also, at the same time, check the box to convert the image to our working space, if we desire.

Adjusting in RGB or CMYK or Lab

In which color mode should we work? My simple, if surprisingly controversial answer is RGB.

Readers of this book who have been exposed to prepress technicians (or who are prepress technicians) will now be frowning and suspicious. For decades, folks in the print trade have told the rest of us to use CMYK files if the destination for our images is print. Indeed, when a commercial printer, like the one who printed this book, asks for CMYK files, that is what I provide. But I first ask them in which color space they should be. That is, which profile should I choose when I convert my RGB files to CMYK.

Bear in mind that a color mode is the general way color is approached: either by adding three colors of light (RGB) or by subtracting light as it reflects off of ink (CMYK). When someone instructs you to “use the CMYK color space,” ask, “which one?”

We need to be specific because a CMYK profile does more than describe the color space of the device to which we’re printing. There are other attributes included, such as dot gain, the paper’s ink limit, and more. So if we were to perform color correction in CMYK, we would need to know at the outset which profile to use. Most of us rarely know on which continent our projects will be printed, let alone on which paper! And in a multimedia world, our images will likely be viewed on screens (RGB devices) as well as in print.

Remaining Healthily Device-Agnostic

Earlier in this chapter, we discussed “Setting Working Space Profiles” (page 215). I mentioned that Adobe RGB was a good choice for those who will likely print their work. It surprises prepress people to learn that Adobe RGB was actually designed with them in mind, and that it neatly encompasses most printable colors.

Photographers who use Lightroom find themselves working in a very large color space called ProPhoto RGB because that is what Lightroom passes along to Photoshop. This color space is large enough to contain all the colors most cameras can capture and seems a good fit for the photo crowd. However, there is no monitor or printer that can produce even close to this range of color. That means what we see on our displays is a conversion from ProPhoto RGB to our monitor’s color space. We may be misled by that view.

Better displays can produce AdobeRGB, and some new, high-end ones can produce the larger P3. Neither is close to ProPhoto. So my considered recommendation for photographers is to use Adobe RGB for images edited in Photoshop. Your originals, likely Raw files, are not affected by this decision and remain in the full range of color your camera provided.

I often set Adobe RGB as my Working Space in all my Adobe apps. I make exceptions for some projects like this book, since it is published as an e-book as well as printed. Because there are quirks in the e-book process, it is far easier to work in sRGB, despite its being a slightly smaller color space.

For those whose images will be displayed exclusively on the web, sRGB is for you. Many desktop printers are standardized to this profile. This is why it is Photoshop’s default.

In any case, you can later configure an image for any print output by converting to that output’s color space profile when you finally know it.

What about Lab?

Over the years, I’ve found myself in dark corners of the internet where self-proclaimed Photoshop gurus claim that mysterious and wonderful adjustments can be achieved exclusively in Lab mode, a device-independent but difficult mode for the uninitiated. Then they spout smoke and hokum to prove their point.

They usually go on about being able to separate color adjustments from tonal adjustments. It is true that if one is in Lab mode, adjusting the L channel has no impact on color, and only impacts tone or luminosity, the L in Lab. But it is also true that trying to adjust color is very difficult in Lab, since the a and b channels are unintuitive, unlike the red, green, or blue ones of RGB. But these gurus brush that off as grumblings of the ignorant.

I’ll be clear: there is no need for us to work in Lab mode. There are ways in RGB to do any adjustment and affect only tone or color (or both) as we choose!

Don’t take my word for it. After we’ve explored the adjustments covered in the rest of this chapter, open an image, save a copy (just in case), and then go to Image > Mode > Lab Color. Try a few adjustments. It might be fun, but I suspect you will not find it very useful.