I've been doing some coloring work recently, I started using layers because that seemed the easiest way of doing, but I found out a few tutorials where all the coloring is done on the RGB channels without the use of a single layer, what is the best method? Supposedly, the later is what pros use.

Which one is better?

I use layers because I don't think I understand channels really well. Apparently channels take up less memory as well.

All my colouring is done on the background layer - I use lots of channels to help with selections, colour holds, and the lineart.

I use additional layers when the initial colouring is done, and I'm completing the image. I'll have a lineart layer, one for the colour holds and trap colour, and another for any special effects.

And I paint exclusively in CMYK. Unless I'm doing a quick goofy IMWAN-style job

Why CMYK Ian?

Most work that I do is for print, and all professional printing is done in CMYK. If you paint in RGB and then convert it to CMYK for printing, there is a colour shift, so the image isn't exactly what you wanted in the first place.

So, even when I'm doing pieces that aren't necessarily for print, I still tend to do them in CMYK just out of habit.

In fact, if you are painting stuff for the web, converting from CMYK to RGB (for .jpgs and .gif's) also creates a colour shift, but a very minor one compared to RGB to CMYK.

I'm a layers girl. Once I discovered them, years ago, I instantly fell in love. Layers have allowed me to create the pieces I have. They give you sooo much more freedom.

This is because the gamut of most RGB color spaces are larger than most CMYK gamuts -- though often CMYK can produce reds and yellows that RGB cannot reproduce, RGB produces blues and greens that are completely out of CMYK's capability to reproduce.

This is because RGB is color based on projected light, based on the (theoretically) 256 shades of grey produced by a monitor, television screen, or scanner, whereas CMYK is based on reflected light and therefore is often limited by the medium on which it is printed, the inks used, and so on.

Guess who does color separation work for a printer for a living?

Oh, and to answer the original post's question, I do everything in layers.

When you get right down to it, I use a mix of both. But not generally manipulating each CMYK channel separately. So, from that standpoint, I suppose you could say that I just use layers. Except it's more complex than that.

The final working file of any image I'm working on will have at least three layers (sometimes up to seven), with at least two channels (quite often more - I've probably used more on that particular image, but tossed them afterward).

The final file I send to a client will just be one layer, though, normally as a .tif.

Thanks, dude - that's the clearest summary of the difference between CMYK and RGB that I've ever read