I often hear the same few questions asked when it comes to mixing with the master in mind. They pop up on forums, I get asked in classes and the answers that I sometimes see are all over the place. So here's what I look for.
Q. How loud should the peaks of my mix be?
A. If you know what you're doing with your mix, I don't care. You can peak at -18dBFS and leave me all the headroom in the world - I'll just end up turning it up. Or you can peak at 0dBFS and leave me no headroom at all - I'll just turn it down.
What matters is that your peaks don't go above 0dBFS and then get clipped off. If I see a mix at 0dBFS then I'm instantly asking the question: has this mix clipped? has this mix been limited to 0dB, or has the mix just come in like that? Similarly, if I see it come in at exactly -0.3, -0.5 or -1dBFS I'm asking: has a limiter been used to get that figure?
Q. Well, in that case how loud should the RMS be?
A. Firstly, it's important to note that the only important figure with respect to your RMS value, should be the difference between RMS and peak. If your RMS is -20dBFS and peaks at -3dBFS then you've got a very healthy 17dB of dynamic range. If your RMS is -20dBFS but peaks at -16dBFS then you've only got 4dB of dynamic range and we're in trouble.
Try and aim for an RMS 12-16dB lower than your peaks. This doesn't mean you can't have louder passages, or quieter ones, but if the average level of the moderate intensity sections is around this region we've got plenty of room to play with.
Q. Should I take the compressor (or other processing) off my mix-buss?
A. If you know what you're doing, then no. That compressor should have been across your mix-buss from early on in your mix, and if you take it off your mix will most likely fall apart. Your fader rides may become exaggerated, your bottom end might turn flabby. The compressor will have placed restrictions on how you mix and in turn you will have made decisions based on that compression.
If on the other hand you just added the compressor to your buss at the end of your mix, to get a little extra level, and because you've seen other people do it. Then, yeh, take it off before you bounce the final mix.
Pretty much the same applies for all other processing, whether it's EQ, saturation or stereo widening. If it's been there from the start, then it stays. Otherwise, think about taking it off. With one exception: Limiters. Limiting a mix can be useful to give a client an idea of what the mix will sound like after mastering, but limiting a mix before mastering ties the mastering engineers hands, and since a good limiter should be mostly transparent, it's best if you bypass it before the final bounce whether or not it was in place from the start or not. This will also help you to see what's really going on in your mix.
Having said all of that though. There's only one global rule about your mix: if it sounds good, it is good.
Any other questions, please ask.