RGB is a method of reproducing colour that is used by light emitting devices, such as your computer monitor. We won’t get too technical with this discussion, but let’s just say that when your image is set up in RGB colour mode, the pixels are assigned values that work nicely with a device that displays colour in RGB mode.
Paper doesn’t emit light, it reflects light. Our eye perceives colour quite differently when viewing an image as a result of reflected light, compared to emitted light. The process by which colour is laid down on paper, generally involves printing just 4 colours, Cyan, Magenta, Yellow and Black – which we refer to as CMYK
If your image is intended for output on a cmyk device (like a printing press), then the pixels should be assigned values that are optimal for that process.
Your image manipulation software (generally Photoshop, not to be confused with your page layout software, namely Indesign) has the ability to save your image in CMYK mode or RGB mode (and a few others, but that’s a whole other story).
Ideally, you should convert your image to cmyk mode before doing any colour correction in Photoshop, otherwise you’re pretty much wasting your time and your work is largely undone when the inevitable conversion to cmyk takes place (whether we do it, you do it, or the RIP on the press does it). Your software will display your image on your monitor differently after it has been converted to cmyk. Because you’re viewing it on a monitor, you’re still seeing it in rgb mode (your monitor is an RGB device, remember), but your software is emulating a CMYK result (printed on paper) as best it can on your monitor. For this reason, you’ll see a difference in how it displays, so that’s why you do your colour correction AFTER converting to cmyk.
If you submit a file that has an rgb image embedded, our rip will simply convert according to standard default settings. Whilst cmyk will give the best results, many people submit RGB images and just allow our rip to do the conversion for them. Ideally you should do the conversion yourself, so that you can fine tune the image after conversion, but if your monitor isn’t calibrated or if you don’t have the experience with colour correction, then you may just choose to leave it be, and hope for the best – not ideal, but probably half of all files we receive are RGB.
If you want to get highly technical with your colour management, there is no better approach than printing a proof, then colour correcting based on what you see in the proof, then creating another proof. As you can imagine, you could end up ordering a lot of proofs, but it’s the most accurate method.