Simply slamming the mix against a limiter is a very rough way of boosting the overall volume. You're on the right track, but you'd get better results with a little more finesse and technique.
I like to compare it to the finishing phase in woodworking; you don't want to just slap the wood across a belt sander, throw a coat of varnish on it and call it done. Instead, you use the belt sander to take out the gross imperfections, then hand sand it with some medium-grit sandpaper, then once more with some fine grain sandpaper. Then stain it, maybe twice, and finally poly coat it for the final finish.
Translate that to audio, and here's roughly what I come up with*:
1. Tame the wild peaks first - i.e. any peaks coming out of the mixdown that are grossly higher than the average peak level. Occasionally there are some wild hairs sticking farther out of the mix than your average signal peak due to noise or simple coincidence in summing. I like to manually edit those down first, though using a limiter to bring them in line can work well also.
2. Run a parametric sweep individually on each side the mix to get rid of "honker" frequencies in the mix. Often times these honkers will reveal themselves in a bad way after substantial compression or limiting, and I find it easier and cleaner to get rid of them now before they stick out.
3. Apply light peak compression to the mix. The exact value varies depending upon the actual program material, but usually something in the 1.5:1 range (give or take), with threshold and attack/release set to work on the peaks above RMS.
4. Listen for and perform any light EQ smoothing that step 3 may have made necessary. This may or may not be necessary at this point.
5. Apply a second coat of light to medium compression, but this time pull the threshold down to maybe 50-70% of the RMS level, to tighten up the whole mix just a bit. Exact values are by ear; if I can hear the compression, I've gone too far and need to back off a bit.
6. For me, some EQ is more often than not desired at this point. It may be either some gentle shaping and nudges at low Q of a dB or two here or there, or it may be a little surgical high Q notching to get rid of harshies and honkers that come out of the woodwork, or it may be some of both. This should yield the final sound I'm looking for. But if I need some heavy EQ at this point just to make a bad-sounding mix sound good, that's an indication that I missed something earlier on and should backtrack to step 2 or 4 and get it right earlier than try to fix it now.
7. Now is when I decide whether to simply boost the output gain to give me my final volume, or whether I still want or need to squeeze even more out of it by running it through a final limiter. If I do limit it, it's strictly by ear; numbers mean nothing to me at this point. There's a point in every mix where it just starts sounding "pushed", and I draw the line there. If I hear that, I stop there and then back off the limiter by a dB or so. That's my final mix. (Rarely does the limiter need to flatten the mix by more than 2 or 3 dB to achieve that result for me.)
*DISCLAIMER: I am an editor and mix engineer more than anything else. While I have done and do some mastering work, I do not claim to be or consider myself a pro mastering engineer such as Massive Master or Masteringhouse or the rest. So take my method with that heavy grain of salt.
Also, this procedure should be considered a general outline, not a fixed recipe. It is an average sampling to give an idea of the general procedure I like to use, and will vary in number of steps and settings used in each step based upon the needs of the mix.
G.
