How do you approach a mix ein general - missing-Basics-question again

my approach to a mix depends on the song itself...for typical rock/metal stuff, i'll take a "bottom up" approach, where i'll start out by getting a drum sound i like, then adding in the bass and makign it play nice with the kick...then i'll throw in the rythym guitars, and tweak everything until it all plays nice together. then i'll throw lead guitars/synths/whatever, tweak a little more...then finally throw the vocals in the mix and fuck with them until it sounds right.

if i'm doing a mix with less "density" where the main focus is probably on the vocal, i'll take a "top down" approach, where i start by doing what i need to do the vocal, then building the rest of the mix around that. if there is no vocal, then i'll build it around whatever lead instrument there is.
 
How long does a mix typically take you, coming from a mix ready session? Mix ready as in you don't have to beat detective, AT, remove noise or anything.
Seems to take me all fucking day to mix something, I'd like to be able to do it in a few hours.
 
What I usually do is that I first remove all inserts, pan everything dead center and bring everything down, so that the set has been reset, then I do the routing. Kick+snare on the first bus, rest of the kit to the second, then bass, then rhythm instruments, then lead instruments, then vocals and then effects. so that makes 6 busses + master.

Then I insert Mr. Mono (for mono compatibility checks) and a limiter to the masterbus set at -2dbFS, just in case I get too excited and don't notice a thing :)

Then the first thing I do is that I bring up the kick and make it sound as good as possible, usually first an EQ, then semifast gate, compression with 4-8ms attack and 5:1 ratio, and if it lacks something, I usually try to add it without using samples. Then I adjust the volumesliders so that the combined sound hits to -6db.

Then I mute the kick and bring up the snare, but unlike others, I usually do the lower mic first. I usually first add the eq, then compression and then saturation & distortion plugs. Then I mute it and do the same setup for the top mic except no distortion here. Then I set the top mic level peaks hit -8db level and then I blend in bottom mic so that it sounds good, which usually is at pretty low levels, like -30..-15db

Then I mix the toms and hihats and same thing here, just the basic eq, gate, compression and saturation again. When the whole kit has been mixed, then I add a highpass to the kit-bus at 80hz and then tweak the the reverb sends and add a lot of compression to the busses too.

When it comes to panning, I usually mix everything from the drummers perspective, so hihats are on the left, ride is on the right etc, but I tend not to use very wide drumpanning, hihats+ride at max -+40%, oh at max+-50%.

For the bass, I usually just record the linesound. I bring in the linebass and duplicate it. I highpass the duplicate at 300hz and add a eq boost at 1khz and distortion plug to it. On the linesound I add a lowpass at 200hz (because there usually is only mud between 200-300hz on bass) and adjust the blend so that it sounds good. Then a highpass at 50hz to the bass bus and a general 3:1 compressor at low level to the bus and blend in with the drums so that they sound balanced. If the kick doesn't get through enough, I don't boost it, but instead I add a sidechain compressor to the bass bus so it ducks. You know, less is more?

Now, as most of you would, I don't mix in the guitars, but instead I put the vocals in now, because they usually (should) sing together well with the bass. If not, there is something wrong with the composition/arrangement. But, some eq, then ALOT of compression, three deessers and another compressor. Who needs dynamics on metalvocals anyways, except for Tool? Then mix in the volume so that it blends in kinda like it seems too loud, because when the guitars kick in, it will most likely be too quiet.

Then I bring up the rhythm guitar panned at dead center, even if it was double/triple/quad tracked. Then I start the EQ:ing only so that it blends in with the mix with the other stuff, no need to EQ if it blends in perfectly, you know? But usually there comes this huge midrange hump that needs to be tamed... When the first guitar is "perfectly fit", then bring up the other guitars and if possible, I leave unnessecary tracks off as it usually keeps the mix more cleaner. When finished highpass the rhythm bus at 100hz.

Then bring up the lead tracks and do the usual EQ stuff again. If it doesnt cut in enough, sidechain compress the rhythm guitar section so that it goes down just enough (no more than 4db compression, as people will notice that!) so that the solo kicks in.

Then go to sleep and come in next day with fresh ears to wonder why this sounds like shit. Then finetune the mix / completely remix the whole thing. When finished and mixed correctly, the whole package should hit peak at ~-4db with ~-20db RMS, which hopefully should be enough headroom for the mastering engineer

How long does a mix typically take you, coming from a mix ready session? Mix ready as in you don't have to beat detective, AT, remove noise or anything.
Seems to take me all fucking day to mix something, I'd like to be able to do it in a few hours.

Two days for me, because I really can't do anything properly in one go. I really have to double check it the next day.
 
Let me bring up this thread again:

Just one question, to get everything right...
Let's say, I start mixing the kick, then bass-guitar, rhythm-guitars etc.
Volumes seem right to me but the master-bus is close to clipping and, for example, the vocals are missing.

I should get everything down a little bit and continue, right? And if I'm close to clipping again, everything else down and add the next track, right?

sounds like you might be starting out a little hot on the faders.

one of the biggest improvements i've made in a long time to my mixing technique is to pay much closer attention to my levels throughout the entire mixing phase. ESPECIALLY at the beginning, because everything builds from there. general rule of thumb is to start the drum mix at around -6db. if that's not loud enough for you, just turn up your monitors. it's seems so obvious, but escapes the best of us. it's really tough for metal-heads (me being the worse :headbang:) to not constantly strive for intensity at all times. this is one time you really need to back off and just let the numbers guide you. the overall volume, presence and intensity will come naturally as your mix progresses (especially mastering at the end).

so why do you want to do this? when i listen to old mixes where i was peeking the meters BEFORE the mastering stage - they sound 'harsh' and, well, crappy to be blunt. most likely because the brick-wall limters and other stuff i had going on were chopping off bits from the top of my cumlative sound stack losing some really important color and nuances of the mix. there was no "breathing" going on - just one big shout. the lesson learned in a few words was: less is more!
 
"Compression first, then EQ" (I don't get it - you would compress everything, including the frequencies you don't want 0_o)

I've noticed that if I EQ -> Comp, the EQ works like a filter "sidechain" and compress all the good frequencies more, making the bad frequencies come back. That's why I use comp -> EQ.
 
I usually start by turning down all the tracks to about -10 using the all group. And turn up my volume knob. I listen to the song while I organize the tracks (sort them by instrument, group and color them). I usually make stereo drum, guitar, and vocal busses next and a master fader.

Then I'll turn down everything to nothing and bring the snare top track up to about -6ish and turn my speakers up until its quite loud, this is going to be about the loudest thing likely so it gives me a reference with some headroom left.

From there I add the rest of the drums looping a part that has all of them being played. I move on to Bass, then guitars, other tracks like keys, then vox, and bg.
I'm adding panning, effects and doing rough automation in the above part.^^

By then its been a few hours and should have taken a break by now.
I keep working at it until I'm exhausted and sick of the song, and then I bounce it and import it into a mastering template I made up.
 
Here's the way I learned a loooong time ago:


Start with the kick, bring it up to about -6db. Listen to the snare and bring it up to be as loud as the kick, don't worry about what it hits on the meters (unless your clipping). Bring up the hats panned half right until they are as loud as the snare. Bring up the toms panned wide until they are a little lower than the snare. Bring up the overheads panned wide until the crash is a little louder than the hats. Bring up the bass until it is as loud as the kicks. Bring up the hard panned guitars until they are as loud as the bass. Bring up the vocals until they are a little bit louder than the guitars. Add reverb to the drums next, solo the snare just enough to make the reverb obvious, then play the whole mix and bring up the reverb just enough to make it obvious on the snare and use a level between those two points for all the drums. Do the same for vox or anything else needing reverb. At this point you should be able to tell what needs EQ because of masking. If you do this once or twice and notice your master fader clipping a little you might want to start with the kick at -9db or so or lower the master fader.


Also once you can do this pretty fast it makes working with a producer a lot easier since you have a set method.
+1 worked perfectly on my last recordings
 
Here's a line.
k, thanks dood.

what?
oh yeah it is to soft. i'll turn it up.
oh shit it's clipping!

yeah, but it sounds cool
***snnnnooooooort***
whoah dude, you just did my line too, what the fuck man?!
oh sorry, i'll pay you back.
alright, turn that guitar up some more
sure thing. whahoh that is a mean ass RIFF.
fuck yeah, dude that's all i write
what?
MEAN ASS FUCKING RIFFS MAN

where are the drums?

oh, they're not loud enough.

k, we'll turn those shits up too.

COMPLIMITCOMPLIMITLIMITCOKECOEKCOEKMMMMFEEELS SOOOGOOOOOD.

it's a hit!
 
I've noticed that if I EQ -> Comp, the EQ works like a filter "sidechain" and compress all the good frequencies more, making the bad frequencies come back. That's why I use comp -> EQ.

depends..


on vox I'd always say even it out with a comp and then use EQ to get the tone.

but sometimes I'm doing it the other way 'round as well (usually not on vox though).
that way you can have more drastic EQ boosts and cuts without making it sound too unnatural.

when I'm high- or lowpassing quite a bit I'd as well use the EQ before the comp...

good example is OH. I use the comp for 2 different purposes on the overheads, one is to reduce snare and kick (than pre- eq) and the other one is to compress the overheads to make them blend more nicely or to make them softer, punchier, whatever..in that case I'm using the comp AFTER the EQ cause I don't want the low freqs that I'm filtering on the OH anyway to drive my comp