Anyway, in various projects since then, several of them comic strips or books, I dealt with these issues concerning RGB and CMYK. I can tell you quite a bit about this.

Obviously the reason a CMYK file is larger than an RGB file is because the CMYK file has four color channels (what we used to call “plates” in printing) and RGB has only three color channels. Each color channel is essentially a grayscale bitmap representing one color element in the image or page. You can actually do some pretty nifty tricks manipulating the color channels in CMYK, but that’s a topic for another time.

CMYK is additive color, while RGB is subtractive color. If you completely saturate a CMYK image you should get black, but in more realistic terms you get the equivalent to the painter’s V. Mucky Pig, which is short for very mucky pigment, a grayish brown. The RGB model works in reverse. The absence of light becomes black, while the combination of all three colors makes white. These two models are different enough to cause some issues in conversion. I should also mention that earlier versions of Photoshop had slightly less effective conversion algorithms than current versions, and there are actually entire software products designed for the purpose of converting from one mode to the other with more accurate and vibrant reproduction. Many professionals spend some time tweaking the color balance of an image after conversion from one mode to the other.

There are two primary models for conversion of RGB to CMYK. I actually used to have a nice big glossy poster devoted to this topic, believe it or not. The two primary models are GCR and UCR. GCR stands for Gray Color Replacement, and UCR stands for Under Color Removal. Neither is considered a perfect process in principle, so conversion in software tends to utilize a combination of both.

Gray Color Replacement is based on the idea that grays in an image should be comprised of a balance of cyan, magenta, and yellow, not just a gray halftone of black ink on the K plate. Under Color Removal is based on the idea that you should try to replace as much of a color with black ink as possible to get the right balance.

I don’t know if you’ve ever experimented with printing at all, but basically if you try to put 100% C, M, Y, and K inks on the page, you get a disgusting smeary mess of ink that won’t make anyone happy. On the other hand, if you use only the black “K” plate for solid black areas, chances are the ink isn’t going to be the expensive, high quality stuff, so you get faded grayish garbage instead of a nice solid black area. Old fashioned comic books sometimes would end up with ugly blacks because of this problem.

The way comics used to be colored was a much trickier process. An artist would color a black and white copy of the art with markers, watercolors, or whatever, and that was called “indications” in the industry. These indications were used as a guide for the person who had to cut film and lay it out to make the correct colors on each plate. They used charts to figure out which shade of gray would be the right one to combine with the other cut areas of film on the page to make the correct colors for each area. It was a huge pain in the ass.

One problem with GCR that could happen was misregistration. You may have noticed the little registration marks on the pages. They sometimes look like little color bars next to a crosshair. The crosshair is where everything is supposed to line up, and the color bars are there to show what each plate’s ink looks like by itself on the paper in use. If the plates aren’t in alignment with each other it’s always a problem. It’s worse with GCR because the black and gray areas on the page are actually a combination of all the colors, so the edges will look like utter crap when they’re out of alignment. With UCR you don’t have that problem because there shouldn’t be any color in your grays or blacks, but it depends on strong black inks to get good saturated blacks.

I always used to wish I could create my own conversion software using special targets for each equivalent color value. It was frustrating working with the defaults, especially when you don’t have the perfect color profile for the conversion. It’s entirely possible in Photoshop for your color profiles to provide seriously inaccurate representations of what your CMYK print will look like, and it uses those profiles for the conversion process. This has led to several disasters in the past which I won’t rehash here.

Anyway, the solution to using CMYK color instead of RGB is to create your own custom CMYK color palette. You can’t just create one in RGB or LAB color and expect it to convert well to CMYK. If you want total control over how your colors print in a comic book, that’s how you do it.

With your own custom palette you can, for example, create two different kinds of black, one that is made only of black ink from the K plate, and the other with a combination of black and color plates.

As I remember it, for some reason Marvel used to want their color work done with the black at 600 dpi as a black and white bitmap (since the linework isn’t shaded grayscale art, so the ink is either there or it’s not) and the color could be done at 300 dpi. They wanted a pure black plate. I doubt they have the same rules now, quite a few years later. They also used to have issues with the colorists dropping the blacks for effect, and I see that being allowed a lot more in today’s comics. I guess they finally got the point all the colorists were trying to make about that.

The main filters you will lose by creating your work in CMYK instead of RGB are the ones that make their calculations based on the behavior of light, such as lens flares. That’s why some don’t work in CMYK. They’re playing with the model for light itself.

It would be nice to be able to create your own color profiles, and there is quite a bit of control available in Photoshop if you want to play with it. You can make it convert from one mode to another in a more satisfying way, but unfortunately it won’t be as user-friendly as an artist would like. Personally I’d like to define the poles of a color gamut for each model. Cyan = this in CMYK, this in RGB, Red = this in CMYK, this in RGB, etc.

Thank you sir.

It also seems like every printer has it's own way of doing things and their own color profile settings. Your best bet would be to ask the printing company you are using of getting the specs straight from them. Sorry I can't be more helpful.