Color Correction Fundamentals: The Basic Tools and Concepts 247 We generally advocate the use of gamma 2.2 for editing spaces, because it's more perceptually uniform than gamma 1.8, and it devotes more bits to the shadows, which is usually where we need them. But there's no getting around the fact that converting a gamma 1.8 image into a gamma 2.2 working space loses some data, though not nearly as much as an actual mode change. Most Win- dows users' legacy images are already in a gamma 2.2 space, so they don't have a problem. Converting legacy gamma 1.8 images into a gamma 2.2 working space can sometimes help if you need to edit the shadows, but it can also make the image fall apart if it has already had heavy edits applied. The one conversion we do strongly advocate is the one that converts high-bit data from your capture device into your working space of choice. RGB vs. CMYK The debate over whether to work in RGB or CMYK has been the subject of countless magazine articles, several online flame wars, and even a few books. Of course, if your work is destined for a film recorder, the computer screen, or videotape, then CMYK is quite irrelevant; but if you're working in the print medium, it's very important indeed. There are still people who continue to maintain that if your work is destined for print, you should work exclusively in CMYK. When confronted, they usually give four reasons for this. · · · · It's the only color space that matters. RGB is meaningless. Monitor calibration is inherently impossible. All that matters are the CMYK dot percentages. While all these points have some validity, we beg to differ with the philosophy as a whole. As we've noted before, when all you have is a hammer, everything starts to look like a nail. People who tell you to do everything in CMYK may have excellent traditional prepress skills and a deep under- standing of process-color printing, but they just don't realize how much image information Photoshop loses during the conversion, and they probably are not comfortable working in RGB. As a result, they convert raw scans (or even worse, JPEGs from consumer digital cameras) to CMYK immediately, damaging the image irretrievably. Then they make huge corrections in CMYK, trying to salvage a printable image from what's left. Once they're done, they congratulate themselves on their exquisite skills while pointing out the limited quality you can attain with desktop color. CMYK Myths One of the reasons we wrote this book was to dispel a number of the myths that have cropped up during the short life of desktop prepress (and some others that have been around even longer)--especially those re- garding CMYK and RGB issues. Here are our answers to two areas that people often find confusing. CMYK has more colors (false).We've heard experts deride the notion that CMYK contains fewer colors than RGB. "Do the math, stupid," they say. "CMYK has 256 4 , or more than 4 billion colors." We wish that were the case. CMYK has more than 4 billion color specifications , but a large number of them are simply alternate ways of specifying the same color using a different balance of black to CMY inks. And many of them (for example, 90C 90M 90Y 100K) are "illegal" specifications that would turn the paper into a soggy mess scattered all over the pressroom floor. When you also take into account the constraints imposed by the black-generation curve and the total ink limit, you end up with far fewer colors than RGB. CMYK is more accurate (true, sort of).Other experts say, "CMYK may have a narrower gamut, but the data points in CMYK are packed much closer together than they are in RGB, so CMYK specifies colors more ac- curately than RGB." Here they have a point. You can specify smaller differences between colors in CMYK than you can in RGB, because the same number of bits is being used to describe a smaller color gamut. (Whether these smaller color differences are detectable by the human eye is a question we'll leave to someone willing to carry out the empirical research.)