I guess I could chip in with some of my experiences.
- Perhaps the single most valuable tool you have at your disposal in your DAW is the bypass button. When mixing a track for a while, bypass all plugins, listen, un-bypass, listen again, repeat and ask yourself the question; "Does this processing land me closer to where I eventuelly want my mix to end up?" Bypass single plugins, hell, even bypass individual bands on your EQ and listen. What does it do to the high-end of your track, how does it affect the perception of the other tracks in your mix, how does it affect the perceived size of your track etc? If the processing takes your mix closer to where you want it to be, keep it. Otherwise, throw it away and start over again.
- One thing I've been trying out lately is what I guess you would call iterative mixing. When you're mixing a track, and you feel like you're done with it, leave it for some time and do something else for a while (either mix some other track or go make some waffles), then come back to it. Duplicate it. Mute one of them and then strip the other one of ALL processing. Then mix it from the ground up again, not trying to think about how you did it before. When you feel like you're done with it, you're left with two basically fully mixed tracks. Compare them. The new one sounds better? Great, throw away the old one and you're left with a marginally better mix! The old one sounds better? Great, throw away the new one and now you have X minutes more mix experience! This might seem like a no-brainer, but it really helped me improve many of the tracks in my mixes, so try it if you have the time!
- Multiband sidechain compressing! Really rolls off your tongue. I do this all the time to improve separation between tracks and overall clarity of the mix. Most of the time, you want elements to sound really full and big when playing by themselves, but might want to roll back certain frequencies when other elements are present to leave some room for them. As an example, it is a quite common practice to automate an EQ band at around 2-4KHz on the rythm guitar bus to go down whenever vocals are playing to make room for them. However, doing this over a whole album's worth of roaring guitars and soaring vocals can be quite redundant. So why not make it an automated process? In this case, simply send the vocals to the rythm guitar. There, receive it with a multiband compressor with sidechain functionality (I use Waves C6 for this purpose). Then make the vocals trigger the compression on one of the bands, so that the guitars duck only in a specific set of frequencies instead of the whole frequency spectrum, which would be the case were we to do this with a regular compressor. Play around with the attack and release settings to make it as smooth as possible. Be careful of overdoing this though, as it will make for a pretty weird pumping effect. If you find yourself with massive amounts of gain reduction, it might be a better idea to instantiate a regular EQ and more permanently cut the frequencies you're attenuating and then try to find a balance between the two.
- Always work the relationship between your EQ's and your compressors. Compressors behave very differently when EQ'd into and you will find your need to EQ might vary depending on how the track is compressed. If you EQ into the compressor(s), EQ the track first, then compress, then go back and tweak the EQ again, play around with the boosts you did or cut some new frequency and see how the compressor(s) respond. Then balance the compressor(s) again these new EQ settings. Working with the relationship between these two can really help you shape the sound in ways impossible when just using one or the other.
- Want your kick to be clickier, your snare snappier or your toms trippier, but have no luck EQ'ing or compressing the track? Try duplicating it, then instantiate a noise gate working at extremes so that only the transient gets by and completely throw away the rest. You might find you need to compress/limit before to get a consistent signal into the noise gate for this to work. Now make that track frikkin' explode, whether it be by EQ'ing/compressing/limiting or even overdrive/distortion, and then blend it in to taste with the original track. Be wary of leakage though when dealing with actually recorded drums, especially with snare tracks, as hi-hat strokes WILL jump up and roundhouse kick you in your face if played at the same time as the snare.
That’s kind of all I can think of for the time being. Pretty lengthy post, but I hope it will eventually help someone out there take their mix a step or two further towards their goal!
