Mastering = what you do on the whole mix after you bounce it to one stereo track.
This is all I do in the mastering stage, my way, not necessarily the best way.
(I do this in
Wavelab 3.04 with VST dynamics and EQ plugins) but you could manage this all in Sonar too.
1. EQ out inaudible areas below 20-25hz and above 22khz (with a smoother curve in the high area to avoid distortion which will happen if the curve is too steep).
2. Compress away 2-10db - I'm producing mostly techno/electronic music so high RMS (leads to lower dynamics) is needed unlike in classical for example.
3. With a maximizing limiter, set peaks to -0.2dB, or limit enough to reduce the most visible peaks with zero attack and sub-20ms release and then normalize. (some CD players clip if you go all the way to zero, or something like that, some pro guy told me some time ago. Not sure if that's true but I prefer to go safe lol)
Now a little pre-mastering tip (before mixing down): you could add a separate EQ to all or some channels, and bring up the most important (lead) tracks. Emphasize a track by giving a peak to it's EQ curve in some area of the spectrum, in the highs for example, and if needed (go by your ear) reduce the same area in some background tracks. Bring contrast.
Hope this helps even a little.
Toni L.
www.mp3.com/NativeAlien