I actually find it helpful to split up the creative process in:
Even if the three can well be thought of as part of the same process, for me it's different mind-sets. Quick breakdown:
- Creating the song not giving a shit about EQ'ing, compressing the dynamics or whatever. I tweak volumes only because I want to hear what I'm creating, not because it's supposed to clear up or whatever. Of course a lot of important decisions are being made during production, and picking/creating the right samples are crucial if you want a good sound at the end of the project. A lot of the "what to pick?" part comes with experience.
- This is when the more technical listening hat comes on. This part of the process is when I take what I got and try not to tweak the song anymore, unless I've planned to add specific effects in the mixing process, f.ex. effect filtering on a group/the whole track or whatever. Other than that, this is where take all the pieces and literally mix them together. Tweaking volumes, applying EQ wherever needed and dynamic compression on separate channels, or possibly groups. It all depends on the project in quiestion. It's not forbidden to add compression/EQ/whatever on the master bus, again this depends.
- When the mixing is done, the mastering prepares the track for the final medium. If it's for a loq-quality web-game, you'll want to master it one way, if it's for cinema another, if it's for "average home listener" a third etc. What can be done differs a lot, but generally you're talking about applying small and well thought/listened parts of copmression, mulitband compression, EQ, perhaps wash the track through some amps to get a small touch of whatever nice and small artifacts that amp ads (which of course can be done in either production or mixing phase too - it all depends on what you're after...)...
...Then again I haven't made music in a long time, and when I have I've only tinkered a little in Reason, so all in all I have no idea what I'm talking about lol