SouthSIDE Glen
independentrecording.net
This is actually a pretty good question. It's also one where not only is a good answer not necessarily all that short, but where there can be a variety of answers, depending upon individual engineer preference; i.e. there's several ways to skin that cat. I'll try to give a fairly generic (but still good) answer, and try not to make it too god awful long (I'll probably fail there
).I think one good way to start the approach to peak management is to take a look at the waveform and roughly eyeball the difference between what are (hopefully) a handful of rogue peaks in the overall waveform. These would be transients that would seem to specifically stick up a few dB higher than the avverage level of the majority of the other peaks in the waveform. Those I personally prefer to manually edit down to the approximate average level of the rest of the "normal" peaks. If there are more than just a handful of them, sometimes just setting a high-gain-reduction compressor or limiter with the threshold set to the average normal peak level to reduce them all in one fell swoop is more efficient and more sensible; I prefer to avoid that when possible, but that's a personal choice only. Either way,the idea is still just to tame the errant peaks to make your peak level more "regular"
Do you do this to the individual tracks or the stereo mixdown? For me, the answer is, "Both". That is, tame the rogue peaks on each track as part of the track preparation when mixing. Then sometimes after mixdown, new rogue peaks can appear because of a coincidence of normal peaks in the individual tracks. I like to knock those down before doing anything else loudness-wise to the 2mix. Do I knock them down in the 2mix or do I go back to the tracks and take them down int here and then re-mix? That depends on what I hear. Sometimes those rogue mixdown peaks are fairly neutral-sounding in character, in which case I'll just take them down right then and there. but sometimes they may actually sound badly by emphasizing the sound of one or more tracks too much just for that split second. In such cases I might go back and take them down in the tracks, but because of the extra work involved, I'll do that only if there's a sonic advantage to doing it that way.
Then once the rogues are taken care of, it's simply a matter of increasing the overall loudness without getting extreme about it. if actually mastering this can be a complicated procedure (and this post is already too long
). But if you want one quick-and-dirty method that can be acceptable maybe 3 out of 5 times or so (IMHO, FWTW) without over-doing the loudness thing, is to take a good quality brick limiter , and when set to a threshold of -0.1dBFS, push the gain until you have a maximum peak reduction of 1.5 dB or so (give or take).HTH,
G.