The RGB model, which computer monitors and TV screens use for display, defines colors in terms of their proportions of each of the three RGB primary colors—red, green, and blue—assigning a value between 0 and 255 to each. For example, pure green has red and blue values of 0 and a green value of 255. To get pure white in the RGB model, you must combine all three RGB primaries at their highest value of 255. In fact, that’s why RGB is called additive color—because adding the three primary colors together produces white. Conversely, black is formed when all three of the RGB primary colors have a value of 0. And when the value of each RGB primary is 128 (half of 255), you get medium gray.

The CMYK model describes colors according to their percentages of cyan, magenta, yellow, and black. These are the four colors of printing inks, both in your inkjet printer and in the high-resolution color printers and printing presses that you see in service bureaus and commercial print shops. A six-color inkjet printer adds light cyan and light magenta to the four basic colors to produce a wider overall range of colors.

Artists usually talk about color by using a set of parameters called HSB, and Photoshop can also use this color model. H stands for Hue, which is the basic color, as shown on a color wheel. It’s expressed in degrees (0°–360°) that correspond to the positions on the color wheel of the various colors. For example, red is at 3 o’clock on the color wheel and is designated as the zero point. Directly across from red on the wheel is cyan, at the 9 o’clock position, with an H value of 180°.

In the HSB model, S stands for Saturation, or the strength of the color, and it’s measured as 100% of the color minus the percentage of gray it contains. Pure color with no gray in it is considered to be 100% saturated. Neutral gray, with no color at all, is 0% saturated. Saturated colors are found at the edge of the color wheel; saturation decreases toward the center of the wheel. If you look at the Apple Color Picker in Figure 4.3, this makes a lot more sense. Finally, the B component in HSB is Brightness, the relative lightness of the color, and it is also measured as a percentage, with 0% signifying black and 100% signifying white.

Figure 4.3. The Apple Color Picker takes the form of a standard color wheel.

The most encompassing of the color models Photoshop can work with is CIE Lab. It defines a color gamut (a range of colors) that is broader than that of any of the other models we’ve looked at so far. Because of its broad color gamut, Photoshop uses the CIE Lab model to convert images from one color model to another—sort of like a universal translation language. Lab colors are defined in terms of luminance plus two components (a and b), which move, respectively, from green to red and from blue to yellow. Lab color is designed to be device-independent, meaning that the range of colors defined in this model isn’t restricted to the range that can be printed or displayed on a particular device. However, you’ll probably never run into a reason to actually use this model.

Let’s start with the most basic of the color modes that Photoshop has to offer: Bitmap and Grayscale. Images in Grayscale mode can contain up to 256 shades of gray, ranging from white to black and covering all the territory in between. Bitmap images use only two colors—black and white (see Figures 4.4 and 4.5 for examples).

Figure 4.4. What we think of as black-and-white photos actually use Photoshop’s Grayscale mode.

Figure 4.5. Here’s how the same image looks in Photoshop’s Bitmap mode.

As you look at Figures 4.4 and 4.5, pay attention to the huge difference in quality. The Grayscale image’s color shade gradually into each other, whereas the Bitmap image’s colors don’t. However, you have a number of ways to convert to image Bitmap mode that can compensate for the fact that it contains only two colors; we discuss those later in this hour.

Whenever you plan for a picture to be printed in black and white or grayscale—for instance, as part of a newsletter or brochure—it makes sense for you to work on that image in Grayscale mode. If you’re starting with a color image, doing the conversion yourself (instead of sending a color photo to the printer) gives you the opportunity to make sure that the picture will print the way you want it to. You can tell by looking at the picture whether the dark parts of the image need to be lightened or the light grays need to be intensified to bring out more detail. You can both adjust the overall level of contrast in the photo and work on individual trouble spots. All this adjustment would take a lot of guesswork if you kept the image in a color mode instead of converting it to Grayscale mode.

The simplest way to convert a color photo to Grayscale is to choose Image, Mode, Grayscale. Photoshop asks you for permission to discard the color information. Click OK to confirm that this is what you want to do; the picture is converted to shades of gray. To convert a color picture to Bitmap mode, as you might want to for certain special effects, you must convert it first to Grayscale mode and then to Bitmap mode; you can’t go directly from a color mode to Bitmap mode.

RGB is the appropriate color mode for working on pictures that will be viewed on a computer or television screen. If the pictures you’re editing in Photoshop will end up as part of a web page, a presentation, or some other medium, stick with RGB for the best color rendition. If your work will eventually be converted to indexed color mode as part of saving it in GIF format, I still recommend doing the color adjustments in RGB and then converting the picture to Indexed Color when it has achieved its final form. That way, you can start with all the colors you need and reduce their number when your work is done. Also, if you work directly in Indexed Color mode, you can’t use layers or any of Photoshop’s filters, and that’s out of the question.

As you saw earlier, you should use CMYK mode only when your image is printed commercially. By converting to CMYK before you send an image to the printer, you can make sure that your nice yellow banana or flower doesn’t end up a muddy brown, or your bright blue sky doesn’t print as purple. Think of it this way: If you’re writing a poem, you want to write it in the language in which it will eventually be published. You don’t want someone else to ineptly translate the poem into a different language, thus changing your meaning and obscuring the beauty of your words. Along the same lines, if your picture will be printed in CMYK, you want to edit it in CMYK.

Desktop inkjet printers are an exception to this rule, however. Yes, they use cyan, magenta, yellow, and black inks, but their software drivers are set up to convert from RGB to CMYK in the most accurate way for their particular hardware. If you send a CMYK image to an inkjet printer, the driver actually converts it to RGB and then back to CMYK before printing it. Imagine what would happen to your poem if it were translated twice in this way, and you’ll get an idea of how inaccurately a desktop inkjet printer can sometimes reproduce a CMYK image.

Indexed color palettes are used to make images smaller by restricting the number of colors they can use. They’re primarily used to produce GIF files for use on the Web, where the time it takes to download an image is always a concern; the smaller the image, the faster it can be transferred from a web server to your computer so that you can view it. Because the colors in an indexed color image are limited, it’s not a good idea to use indexed color for continuous-tone images (such as photographs), because they use a lot more colors than you might realize as each color shades gradually into the next. Indexed color is great, on the other hand, for logos, simple illustrations and charts, and clip art because these images generally use fewer colors than photos in the first place.

Photoshop offers you several ways to build a color palette when you convert an image into Indexed Color mode. You can choose how many colors the palette contains and determine how Photoshop chooses those colors, or you can even pick them all out yourself. From RGB mode, choose Image, Mode, Indexed Color to take a look at the Indexed Color dialog box (see Figure 4.6).

Figure 4.6. The Indexed Color dialog box offers you several ways to convert an image.

To reproduce colors that aren’t included in their restricted palettes, indexed color images use a process called dithering, in which adjacent pixels of different colors are interspersed, visually blending onscreen to simulate a new color. If you zoom in on a dithered image, you can still see the pixels’ original colors—or the closest index equivalent.

Your palette choices when converting to Indexed Color mode are as follows:

All you have to do to convert an image from one color mode to another is choose Image, Mode and then choose the mode you want to use. Now, keep in mind that because different color models have different gamuts, Photoshop might have to change some of the colors in the image to fit it inside its new gamut. Photoshop uses the Lab color model, which has the broadest gamut and contains all the colors that other modes use, as a way to convert images between color modes. But this is no guarantee that your colors will look the same in another mode as they did in the original mode.

I’ll bet you can guess what I’m going to say you should do to combat this fact of life: Do your work in RGB, even if you plan to have your images professionally printed. Convert a copy of your image to CMYK just before you send it to the print shop. First, though, use the Gamut Warning command (described earlier in this hour) to see whether all your colors are within the CMYK gamut. If you are going to publish your images on the Web, stick with RGB or use Indexed Color mode for files that you plan to save in GIF or PNG-8 format. Following these guidelines will save you many hours of wondering why the web page logo that looked great on the office Macintosh looks funky on the Windows machine you use at home, or why the yellow in your printed piece has a brownish cast.

If you have a color printer, try reverting to RGB, and then print your picture and compare it to what you see onscreen. Does it look OK? If so, you’re in luck—your monitor is accurately calibrated. If not, you need to calibrate your monitor so that the images onscreen accurately display the colors as they will print. Calibration is covered in the “What’s Color Management?” section in Hour 23, “Printing and Publishing Your Images.” If your monitor seems to need calibration, you can jump ahead to that section.

It’s actually a little bit misleading to call 8-bit color by that name: What you really get with 8-bit color is 8 bits per color channel. So depending on the color mode you’ve chosen to work in, you have anywhere from 8 to 32 bits of information for each pixel. Because grayscale images have only one color channel, an 8-bit grayscale image really does have only 8 total bits of color data. In RGB mode, you have 8 bits each for the red, blue, and green color channels, making 24 total bits of color information. And in CMYK mode, you have 8 bits each for four channels, or 32 bits in all.

In older versions of Photoshop, not all of the tools and filters worked with 16-bit color. In Photoshop CS3, all tools and most filters became 16-bit compatible. Does this mean you should do all your work in 16-bit color? No. Technology moves forward in uneven leaps. Even though Photoshop can work in 16-bit color, your monitor can’t show you all the millions of colors your image can contain at the higher bit depth. And your home/office inkjet or color laser printer can’t possibly reproduce them. This could change eventually as new HDTV monitors and better printing inks come on the market. The next frontier will be 32-bit color—but Photoshop’s support for that at this point is still so limited that it’s best to leave it alone.

You can find out what kind of color an image uses by checking the setting in the Image menu’s Mode submenu. If the checkmark is next to 8 Bits/Channel, for example, then the image is in 8-bit color. Converting upward to 16-bit color won’t make a difference, so the only time you should see a checkmark next to 16 Bits/Channel is when the image came to you that way.

Here’s a chance to get your own mental picture of which RGB colors just don’t translate to CMYK. First, create a new blank document in RGB mode. Choose View, Swatches to bring up the Swatches panel; then switch to the Paint Bucket tool. (It’s under the Gradient tool.) Click the first red swatch at the top of the panel and then click in the window to fill the image with red. Now choose Image, Mode, CMYK to switch to CMYK mode. See how the color darkens? Press Command-Z/Ctrl+Z to undo, and then fill the window with the first blue swatch at the top of the window (not the cyan one—we want royal blue, which is the blue from RGB). Switch to CMYK mode again. See the huge color change? These two colors, true red and true blue, just can’t be reproduced exactly in CMYK. Now repeat this process with different colors from the Swatches panel, each time reverting to RGB mode before filling the image window. Note which colors change the most when you switch to CMYK and which ones change the least.