depends on what purpose the track is to serve in the mix.
I tend to over-compress at first, in a hunt and peck fashion, in order to get the mix sitting reasonably well, level wise. then go back and re-examine what you actually did to get the levels working together. This will be an examination of multiple tracks >>a combination of level cuts or adds at the beginning of the chain (from View or EQ screens), EQ adds/cuts, dynamics ratio/threshold/knee, dynamics makeup gain, and fader position. Then I'd work on individual tracks, releasing (usually) compression by backing off ratios/thresholds etc., listening for the mix to "open up" as the compressors get relaxed, but not losing the levels established in the first step. A very iterative process. I make what changes are necessary along the chain in order to have the fader sitting at about 0 dB
As far as how hard to make them work?? My benckmark is 3 db of attenuation. But there is the interplay of attack/release times to consider when judging "how hard" the compressor is working too.
Less is usually better, but ....
In more complex mixes some tracks I compress rather heavily and provide with generous makeup gain, but rather low fader levels in the mix. Such tracks are sometimes useful to sit below/behind the mix and act as foundation. These may probably be attenuated more severely than "featured" tracks.
Then when the mix is "almost complete" - go back and look at the compression levels of the main tracks again for final adjustment ( release or tighten to preference).
Mild compression on the stereo track is another "tightening" strategy. So don't be afraid to loosen up the mix as it develops, using the approach i have described, and then experiment with GENTLE very fast attack/quick release compression on the stereo track. The track can then go through one "final mastering" step
IMHO the final step is best accomplished away from the AW. The AW does not seem to be able to provide a decent brickwall limiter. This is essential to lifting the mix's overall and average RMS volume level, to "commercial" levels. I would prefer to have a very gently working brickwall limiter on a computer editor, than to struggle with this aspect using the AW's dynamics offering. I use my y96k card to accomplish this final, vital step within the machine, which is why I like the card so much.
Lots of ways to skin a cat.
_________________ Byron
|