I read Paul Giangregorio's tutorial regarding Sonar. I would like to offer an alternative to his mixing style. I would be happy to hear comments on the differences.
Paul seems to do all his mixing on the midi side. I do not. Once I've finished scoring, I fill up tracks with sound (without going over 0 db) and convert all tracks to audio. I do only one midi to audio conversion. Once tracks are converted to audio, I archive all midi tracks. I apply effects and mix totally in audio.
I use a bus structure similar to Paul's tutorial, but I find one must apply some effects to individual audio tracks in addition to buses. Once I tweak the individual tracks, I use the freeze function on individual tracks and send them to buses, where I apply additional effects, pan, balance, etc.
A few techniques I use that might (or might not) help in creating gorgeous sound: 1) I don't use the mod wheel while recording. I fill up the track with sound, then use automation envelopes after I convert to audio to do everything the mod wheel does. 2) I layer midi tracks (and there are effects I do to layered tracks different from the dominant track) and place them under the dominant track, and 3) boost the high end EQ on the master just a little bit in my final mix.
Paul says folks should spend most of their time writing and less time mixing.
I spend as much - or more - time mixing and mastering as I consider those functions to be highly artistic.