Also, random trivia - did you know Winamp seems to have a built-in compressor algorithm?
I was sitting here listening to "Caught in a Web" when I first queued it up and thinking, "Man, they squashed the ****
out of this in mastering, I don't recall it sounding nearly that bad," only to notice I'd accidently clicked the 3khz slider up to a full 12db boost. This raises the output way above -0db, and it sounds like Winamp was able to detect that and limited it on playback to adjust.
Wow, Drew, that last post was more clear more than the 3 months of research! that makes a ton more sense, its just basic math i see lol!! wow, I feel kinda retarded now lol, but at least it makes sense!
Yes and no. Long story short I've been dabbling/getting increasingly serious in home recording for a bit over a decade now, and there's a couple things that have become increasingly clear to me the more I do it. Chief amongst them is gain staging is everything.
So, sometimes it IS just math. Digital "gain" is pure right up to the point where it clips- if you take a .wav file and up the volume by 5db, then provided that extra 5db won't push you over 0 you'll get perfectly pristine added volume - no new noise will be added to the signal, though of course any existing noise will come up 5db as well. If you then, also in the digital realm, take that same signal and knock it down 5db in your DAW, then again it's pure, pristine gain (negative, this time), and all that happens is the same exact signal plays back 5db quieter than it was. So, if you add two gain stages in your DAW, one increasing it by 5db and another decreasing it by 5db, the net result is precisely zero.
However, for a gain stage that can potentially color or change the sound, that's NOT true. Analog gain stages such as those in the preamp of your recording interface definitely fall in this category - the difference between recording with your signal peaking at -0.5db and recording peaking at -10db and then using a volume plugin to boost it up to -0.5db is potentially huge. Essentially, at input you have two concerns - one, getting as good a signal-to-noise ratio as possible, keeping in mind that as you increase preamp gain you can potentially be increasing noise, and two, to make sure that whatever else you do, you're not screwing with the transients on your peaks, that they're being reproduced cleanly and crisply as possible. It's tough to gauge this by listening, especially to a track solo'd, but as you start to layer track after track with less-than-perfect transient response the result gets increasingly murky and muddy and dense. I can't offer you any more than anecdotal evidence here, but after reading a few good articles on this stuff I stopped trying to track between -3 and -6db (and then having to turn everything down in the mix) and started shooting for -12-18db (and have to do no more than balance the tracks against each other, boosting one or cutting another here and there, and have a finished mix peaking nicely around -6-8db), and immediately started getting clearer sounding mixes.
All of that is pretty esoteric, though - if there's anything you should take away is that given the headroom 24-bit allows, if you're finding you have to turn all your tracks down in the mix to keep your master bus from clipping, you're probably tracking too hot.
That said, the single easiest (and coolest) way to demonstrate how much gain staging matters in a mix is this - open a track, and load a compressor and an EQ. Put the EQ after the compressor, and come up with some truly stupid settings - say, a 24db boost at 3k with a narrow Q, and then a 20:1 compressor with a threshold somewhere around the body of your track. Something absolutely dumb
that you'd never use unless you were trying to make something sound unnatural. Now, hit playback and listen for a while. Then, while playing back, switch the order of the plugins, so instead of Compressor => EQ you're going EQ => Compressor. What before sounded like a kind of unnatural guitar suddenly sounds like an even thinner version of that guitar that's absolutely getting squashed
. The EQ plugin is adding a lot of gain to your signal - 24db, in fact, in a very narrow band. When it's happening after the compressor, then the compressor doesn't "hear" that extra gain, so you compress your signal and then boost it in the EQ. When you flip the order, however, suddenly the compressor is seeing this huge raging 24db spike that it has to clamp down on. The effect is twofold - first, it makes the compressor work much harder and more audibly, and second it lesens the impact of your boost - it only comes through as a couple db higher there because the compressor is slamming anything that passes its threshold, and your boost is way over that threshold. IT sounds aweful, and frankly is kind of unpleasent to listen to because your ears aren't used to hearing something that flat.