In my most recent song, I feel like I really nailed the mixing, and I wanted to write up what I did. I posted #3 on the forums, but I felt like I could expand on my process here. Here's the song.
http://www.newgrounds.com/audio/listen/622648
How did I do it?
It was not due to 1 weird trick that audio engineers don't want you to know ;-) Rather, it was the accumulation of a bunch of small things.
1. High pass everything
Think of the frequencies in your song like an inverted pyramid, with the point of the pyramid at the bass frequencies. The lower you go on the pyramid, the fewer elements you want to be in the mix. This is why things like delay and reverb are heard mostly in the higher frequencies, and people often advise to cut out lower end of reverb, because it sounds 'muddy'. Too many things in low frequencies will make the song sound like crap. Too many things in upper frequencies will make the song sound interesting. (To a point. Do not go overboard with this advice. :)
Seriously, listen to your favorite EDM track and count the number of things you can hear in the upper range. There's probably hi hats, high ends of almost every instrument and percussion element, delay, reverb, random effects, white noise... it goes on and on. Now count the things you hear in the low end. There's bass and a kick. Not even that, but you have to sidechain the bass to the kick so that they don't overlap! (Imagine if you had to sidechain your high-end elements together. You would die before you finished a song. ;)
Alright - hopefully that was enough to convince you. But maybe you're thinking that you don't put too much stuff in the bassline, but it still sounds muddy. You might be surprised to find that almost every instrument you use has low end noise that is going to get in the way of your bass and kick and muddy up the low end (remember! only 1 thing down there!). If you don't believe me, open up your DAW and attach a spectrogram to one of the instruments and start playing around on it. You'll be surprised to see all the crap that it has going on in the low end. It's fine to have that going when the instrument is just playing by itself - it makes it sound more 'full'. But when it's playing together with your bassline, it's going to clash and sound terrible, so you should high pass it to cut out the frequencies you don't need.
2. Make your drums louder
The #1 easiest way to spot a mix by an amateur musician (not to diss on amateurs or anything! I still am one!) is to listen to how loud the drums are. Now, if the track is not supposed to be energetic, disregard this advice, but otherwise: I can guarantee that almost every beginner musician will make their drums far, far too quiet. In an energetic song, drums should be the loudest thing in the track.
The reason that everyone gets this wrong is because no one listening to a track is paying attention to the drums! With the exception of the odd drum solo, your attention is always on the lead or the vocals. In the same way that you may not be fully conscious about the fact that it's actually the bassline of a song that gives it its fullness and weight, you might also not realize that the drums are what are giving your favorite track so much energy. When you think about it this way, drums and bassline are kind of the unsung heroes of music... (except in dnb)
Even though they're the loudest thing in a track, you probably aren't even consciously aware of them. So here's a homework assignment: Go listen to your favorite EDM song and pay attention to how loud the drums are. Maybe you can't tell how loud they are relative to the rest of the track. (I have difficulty with this as well, just because it's so hard to break that subconscious processing.) In that case, try dragging the mp3 into your DAW and opening up a spectrogram and looking for spikes when the drums are hiting.
They're pretty loud, eh?
Alright, let's say you're convinced now. So what do you do? Well, obviously, you can just crank the gain on those suckers. Turn down the rest of your song if you have to, and make them the loudest thing in your track. Boom, you're practically done.
The other thing you can do is compress them. A guy much smarter than me told me that you should pretty much always compress your drums. It's essentially a way to get them to be even louder, still without making your song go into the red.
Yes, people really like loud drums that much.
3. Stereo Widening
I think that one of the big successes I had in my track "Light Up" was mixing the pad rhythm to fit well with the melody line, even though they're both at the exact same frequency and should therefore technically overlap and sound terrible. Listen to e.g. 1:20. If you go and try to replicate this in your DAW, you'll almost certainly be aggravated. The pads will stomp on the melody, or they'll be unhearable.
The way that I did it was with <em>stereo widening.</em> Essentially, Ableton has this neat tool that allows you to take audio and split into two copies of itself, then pan one hard left and one hard right. If you use Ableton, it's the "Width" option in the utility. If you don't use Ableton, I believe you can replicate this by duplicating your audio channel and panning one copy to the right and one to the left by equal amounts. In Ableton, I made the width 150% (100% is default), which is probably equivalent to panning the 2 channels halfway to left or right.
This will separate the pad from the rest of the song (which is right in the middle, I'd assume). Not only will it allow the pad to live harmoniously with the melody, but it'll also make the song sound fuller because you're taking advantage of the full range from L to R.
Stereo widening is awesome.
Alright. I have even more tips I could write here, but I think that 3 is a good start for now.
