PART 5: MIXING
"This is art, not mathematics. There is no "wrong." Jazz session drummer, Kevin Schmidt.
The best way to teach is by example. I can ramble on for days about things, but to help you truly understand, I've put together a very short demonstration.
Here's some tracks, first raw, then mixed, then with accompanying instruments:
You can grab the file
HERE
"Just Do It. ...Later." --T-Shirt I saw earlier today.
Ok, I know this installment is seriously overdue. When I started this thing I was starting five week layoff from my day job & I had a grand total of one project on the go, and it was only a 2 song deal. Here we are some eight(!) months later... I've got 2 albums on the go & one more to start before year's end... not to mention I'm back full time at my day job.
Not only that, but I've seriously been wrestling with the notion of how to write this segment. Mixing can be a difficult task, and how to put my ideas into words can be quite challenging. Chalk it up to writer's block. I know what I want to say, but it's been difficult to say it. So, I'm gonna take Strongbad's advice here: "Hold out your hand like you're holding a giant mystical orb and let the evil flow...."
Philosophy
So, you've got your record tracked. Playback is pretty cool. Guitars are thrashing away & your vocalist sounds like he's gargling razorblades. But the snare is a little dull. So, I guess you should reach over and crank up the 10K on the snare mic to brighten it up, right?
FUCK NO!!
Ok, repeat after me: "Boosting EQ sucks." Why? Well, that subject could be a book, but to simplify: When you boost an EQ, you're adding noise. And on a snare mic, it's like a "turn up the hi-hat knob." Which is what we don't want.... that is unless you're doing disco. If you have a wood floor, you won't have to reach for a boosting EQ nearly as much, either. Subtractive is the way to go here, folks. The end results will be cleaner.
Don't get me wrong, both have thier place. But a good rule of thumb should be "Cut first, boost later."
Some tools you'll need: A compressor plugin with a backend limiter on the makeup gain, & a DAW that can sidechain. If your DAW can't sidechain, you can get a copy of SawBasic for $300 US, and it's faster than anything you've ever seen.
That being said, I'm still very much an old school guy. I try to avoid drum replacement as much as possible. I'm into capturing the sound of the band, as opposed to the band having my stamp on them, and having the mics work with each other. So, where do I think drum replacement is appropriate? Well, let's start at the bottom:
Kick Drum
After you've spent so much time adding new skins, falam slam pads, changing out beater types, moving mic positions & generally tweaking to perfection, this is the drum, for me anyway, that most likely winds up being replaced. Why? Dynamics. Some drummers just naturally hit a little softer during fast double bass parts... and look out if he doesn't hear it in the mix. Now, there's numerous ways to deal with this. Gating, compression, etc, etc, but the harder you compress the kick mic, the more background noise you bring into the mix. That means cymbal wash, and more importantly, bottom snare noise.
Generally, if the drummer has put in a very solid performance on the kicks, I'll stick with the natural mic. But if not, I'll have him play some solo kicks at the end of the day & make up a custom .gog file for the record. It's really not that hard to do, it just takes some patience as there's lots of cutting & pasting the hits into individual files. If you've tuned your drum right, have a good mic & pre on it & the kick sounds wonderful, you'll have a pretty killer sample to work with.
I find the kick is the easiest part of the kit to replace. Why? Because the beater only hits in one spot. The only variable is dynamic strength... position doesn't matter. If the drummer is using two bass drums, you can sample each one & run a pair of drumagogs for that authentic 2 bass drum sound. A little bit of variation from drum to drum is a good thing. This avoids the "machine gun effect."
Replacing the kick also give you room for some boosting EQ to bring out that all important "click" sound. Since it's an "after the fact" sound, you don't have to worry about a big boost fucking with the phase of the other mics.
Usually, I'll put a pretty wide scoop in around 500hz, and a boost around 8-10k depending on the sound I'm after. That, and lots of compression. 15ms or so attack, and around 100ms release, with plenty of makeup gain. If you're using drumagog, one critical point: Remember to click on the "advanced" tab & pull back the "dynamic tracking" knob. This will really even out the kicks. Here's a shot:
Pull back the dynamic tracking knob to even out the kick hits
On the other side of the coin, with a natural kick, I'll sidechain thru Saw's channel strips. In the pic below, I've got the key coming in from channel 2, which is the snare. Attack 0, release 20, a gentle ratio & a very low threshold. This will step on any snare drum transients bleeding into the kick mic.
But if you'll notice, I've got 2 kick tracks. The top is using drumagaog, for the attack. The second track is a copy of the first, but it's the live mic, with a lowpass filter set around 200hz, and a bass enhancement plugin. Mix the two together & you get a pretty slammin' kick!
SNARE
This can vary greatly. Each drummer has his own style of hitting the snare, from strength to position, and it's your job to make the dumb shit sound good. One HIGHLY, highly recommended tool is the
Dominion plugin from digital fishphones. It's free, so why don't you head on over to
Dominion Download Site and grab a copy.
Now, if you've lambasted the drummer enough about the hats, hopefully he's backed off enough & you've got something useable. Hopefully you've also put the mic in the right place & minimized bleed.
If not there's a few tricks to get around it. First, you can gate the snare. With an ultra-fast attack & a sensible release, you can get away with a lot. The other idea might be to use an expander. This will drop the noise floor, i.e. the goddamn hihat bleed, in relation to the snare. Dominion has one built right in, and it's great.
So, EQ is next. As you can see, I've violated my own rule & used a pretty healthy boost around 8k on this snare. I was able to get away with this because the bleed on the snare mic was pretty minimal. If the hat was louder I'd have to get a little more crafty. Physical separation, mic placement, and most importantly, a light touch on the hat make life easier for the engineer.
I've also got a 1.5 dB cut at 460 & a 2.5 db cut at 100, and the mic is rolled off at 75hz. This will vary from drum to drum & drummer to drummer, so don't take these numbers as concrete rules.
I'm also using Dominion, post channel EQ, to drop the bleed. Look at the "sustain" faders and you'll get what I mean. Dominion also has a pretty cool high frequency exciter, under the "hf details" knob. This can be a very cool alternative to boosting EQ. The attack faders are used for tuning the incoming transient. Overall, I've found this plugin to be indispensible. You can tweak the attack & release of a drum, drop the bleed by several orders of magnitude, and generally fuck with sounds till your hearts' content. On this particular track, I felt the snare didn't need any more attack, just a little bleed management.
After dominion, I'll use a plugin comp to bring the level waaay up. If you've got one with a backend limiter, say, a Sonditus comp, this will be your best friend. Set for about 15 ms attack, & crank it up till the output limiter light goes off. Say, 8-10 dbs. Ba-Bang!
OVERHEADS
This has gotten a lot easier since I moved over to spaced pairs. Rolloff at around 600 hz, depending on the drummer, (Mr. Sneap's idea. Just fn' brilliant!) and sidechain from the snare. Attack 0, release 20ms, ratio & threshold to taste. This will take the snare transient out of the overheads & leave some of the sustain. If you've set your threshold right, you can get a great effect with ghosting on snare rolls. It will leave the softer hits in the overheads & only step down on the hard ones. The rolloff with clean your mix up big time as well. Cool!
The other nice thing about sidechaining from the snare is it turns off on the tom rolls. This is great at it adds some pretty heavy attack to the toms in the full mix. Great stuff.
After the sidechain, I'll usually use a plugin comp to even out some of the cymbal hits. Theshold at around -15 to -18, 4:1 ratio & perhaps a bit of makeup gain. Put this up with the snare & things should be sounding pretty explosive.
Toms
Pretty much the same idea as the snare. Mid scoop, & smashed with a comp. Be sure to rolloff the lows though, so your speakers don't puke on the floor. Dominion can be a great help for adding attack if your drummer can't hit toms very hard. That & the 'hf details' knob can be a big help too.
I'll usually set up SAW to auto-remove silence for the tom tracks, & remove the bleed when the toms aren't being played. Saw has a pretty cool feature to add a 'reverse' to the gate when removing silence. What this means, is that if you set it for say, 5ms, it will open the gate 5ms BEFORE the tom strike, thereby letting the tom thru with all it's attack. Man, what a timesaver. Generally, after silence has been stripped, I'll listen to every fill & make sure things turned out correctly.
REVERB
Don't overdo it. Seriously. Pick ONE reverb you like and stick with it for the song. I set the verb's tail to the tempo of the song, so it's not stepping over itself. Generally, I'll put the ver on a return buss & take a send from the snare & toms to a verb plugin, & that's it. Kick? Almost never. Only in very special cases, when it's all out in front by itself. Overheads? Even less. It just clouds up the mix.
One critical point: On your verb return buss, add an EQ plugin before the reverb & rolloff everything below 200hz. That way, you're not reverberating the mud, only the mids & hi end stuff. This will clean up your mix bigtime.
If you listen to the mixed drums in the .rar file, you can hear the only thing with verb on it is the snare. Since it was an old school tune, we used a fairly large verb. But, the key is to go with what works for the song. Talk to the band you're working with and make sure you can get what they're after. These days, a big fat verb on the snare is usually not what people want. It usually makes people cringe "Bon Jovi." The trick is to use 'just enough.' Because even a short verb can greatly enhance a snare sound. Same with the toms. Make them sound like they're on the inside of an oil tanker only when the song calls for it. The rest of the time, make sure there's 'just enough.'
BTW, here's a link to the full song:
Final Stage - In the Night
Well, that's it for now. It's a pretty basic mix guide, but hopefully someone will find something useful in here. Feel free with the questions & comments.
-0z-
