Mix Report – 5/10/15

And here we are.  Another week has past; the mix is still in progress.  I was able to log somewhere between 8 and ten hours last week.  Not bad, considering my day job.

The week, from the mix perspective, started off bad.  I was revisiting a few songs as my abilities (my ear) had improved since I last worked on them.  One night was spent revisiting four songs.  Well, revisiting three songs, and a tweak to one which is… you know… almost there.

Free, The C is Silent,  and Lunchroom/Atrium are feeling pretty done (again.  I brought Someone Ran Down the Road Screaming, Self Worth, and The Accident back into the fold.  

Re-approaching those mixes on one night was like bombarding myself with misery.  After a couple of hours, I felt as though they were all worse.  It happens; every so often the studio defeats me.  Pretty common feelings involve:

  • My mixes are shit
  • I can’t finish this
  • I’m in over my head
  • I’m spinning in circles

Truth.  It’s part of what I get to deal with on top of doing the actual work.  Part of what is tricky is, I’ve never finished anything before.  I don’t have much of a precedent.

Except I do.  Three songs are feeling really good.  I’m sure I can get the rest there.  When I have these nights I remember to step away, and come back again… and again… and again.  Sooner or later, I get there.

Someone Ran Down the Roat Screaming was my song of choice.  Early on in the mix, I thought this was my best mix.  Now I see it as ‘needing a lot of work’.  I spent a couple of nights fighting this mix and getting no where.   Nothing I did was helping.  There was a viel of shit on the mix, and I couldn’t seem to help it.  Literally, nothing worked, or the changes I was expecting to see from what I was doing, felt so minor.

But, as I do, I learned something.

The Viel of Shit

I’m starting to categorize different types of mix problems.  Some are basic seating ones (more on this below).  Other problems in your mix, fuck your whole mix.  I mean, a single mistake can pull this sonic curtain over you and you will get fucking no where.

In this case I decided I would try to do something I knew was dangerous early on in the mix.  Early on it sounded great (I was probably high), but in this application I had gone too far.  

What was I trying to do?  Multiple reverbs for fatness.  What does this mean and why was I trying to do it?

Basically, I think of reverbs of having one of two types of  applications.  One is getting things to gel (or sit) in the mix.  The second is for an effect on a single sound.  For example, sitting in the mix – that means I want everything to sound like it’s in the same room, so I apply the same room sound to all sounds.  The second, the effect, maybe I want a sound to be really big such that it adds extra interest to it.

I wondered if I could have both of these.  I thought about a rock concert, and it made some sense.  In the rock concert you’re in a room (thus the room is applied to all sounds), and it’s not unheard of for a guitarist to have a reverb affect on himself so that he sounds nice and big.  This is what I was trying to do with plugins.

If I take a hall and apply it to all my instruments to get them to gel, what if I send a couple of instruments, just a little bit, to some epic cathedral.  This is exactly what I tried to do.  

Where did I go wrong?  If a busy sound gets too much of this, it’s verbs will over power the mix and you won’t hear past any of the intense mud destroying everything.  I had a sound at -45 on a send, this is way too loud for heavy verb application like this  (the other verb’s are all -15-20 range-ish).  This technique really breaks down when I start pushing past -60; I can sort of get away with -55 or even -50 on really sparse instruments.  I find it provides an interesting yet subtle fatness… but too much made the mix unwieldy.

So that, was lesson one.  Once I found that one extra reverb signal that was too loud, it was as though I could hear all kinds of things.

The Seating Cycle and Convergence

I also wanted to share some ideas that I’ve been toying with for a bit, particularly in the last couple weeks.  When you talk to a mix engineer, often times you’ll hear them say things like “Man, I wish I could get those vocals to sit in the mix better.”  

Taking a seat, so to speak, is an interesting concept.  Particularly interesting for me to try to explain (maybe I’ll get it wrong and incur the rath of mix engineers everywhere).  At the end of the day, in your mix, you want to be able to clearly hear all the instruments (more or less), and have a general frequency balance such that everything sounds natural (if using analog instruments humans have a basis of comparison), isn’t fatiging or jarring in a strange way, and that translates to as many listening devices as possible.

Given that run on sentence, I see the ‘seat’ of a sound is where not only are these balances perfect, but the sounds nuzzle up to each other like a cuddle puddle at your favorite rave.  Every sound have a place, it snuggles up to the sounds next to it, but every sound has just enough space.  It’s very meta, but your ear knows where to go.

I’ve found, for me, in the persuit of ‘the seat’ I have a bit of a cyclical process.   It spins round and round, until sooner or later everything lands, just right (converges).  Before I get into the cycle, I should say that this pairs up well with the tapestry thing I talked about in my last blog (that all the sounds are connected).  So as you fight for the seat in one sound, you may end up re-seating another sound in a very similar way.

So, the process (and it’s complicated and ad-hoc, but I would like to try to describe it.

  • Fader (level of signal)
  • Reverb (The space the signal exists in)
  • EQ
  • Rinse, repeat, and play with other sounds too.

Faders, Reverbs, and EQ’s can all lend to both physical notion of where the sound is (front or back), and of course, how the frequencies play with each other.  I’m probably not getting anywhere, but consider for a second a thought process:

  • Crap, these hats are too hot (too in front)
  • I’ll bring the fader down a db
  • Now they are too behind.  Hrm
  • I can back the reverb send off a half db
  • Now it’s almost there, but still a little weak
  • I’ll bump 3Khz .5db
  • Hrm I’m have a new clash with a snare
  • Hrm, what next

But, as I look for this delicate center for the sound (a hat, or whatever), I’m spending tons of time with my levels, my reverbs, and my eqs.  Almost exclusively.  And, each change I make, affects how that sound relates with other sounds (tapestry theory).  

I spin round and round, but slowly, nudging one sound after another, revisiting each sound to seat it and re-seat it… sooner or later, there become less and less of these changes to make.  

The sounds start to converge on the same seat.  At least, if I don’t spend my time making the same two changes over and over again.  You learn about your sounds and how they relate as you do the mix; I think understanding these relationships is part of what helps you find this point of sonic convergence.

Did I lose you?  Ha, it happens.  But know this, I’m starting to see the convergence for Someone Ran Down the Road Screaming and it’s brings with it a sense of relief.  The notion that, “ah ha, I might just be able to finish this mix after all.”

Convergence and Fine Tuning in Ableton

The closer I get to finding that perfect seat for all the sounds, the smaller my changes become.  Early on I make big changes in my faders.  Toward the end, 3db, 2db, 1db, and even sub-db changes looking for the perfect place for each setting.  

Since I’ve been in this position so rarely, and didn’t have the equipment, or the sonic know how to tell, I never really thought about how to do this in a DAW before.  ProTools has the notion of nudging (at least, in reference to sounds and time).  Logic has this too.  But what about subtle changes in automation or to a setting?

Most DAW’s I’ve used you can hold down a modifier key (ctrl or apple or whatever), and drag with a very slow change to whatever setting.  Other apps do this too (non-audio).  

Sadly, I couldn’t find a key that did this in ableton live.  Searching around regularly yielded nothing on how to do subtle changes to a fader or automation.  Either ableton users just don’t do this stuff, or I’m missin something.

I found that, when plaing with session view, if I expanded a channel strip a little bit I would get additional decimal places on my fader settings.  15Db became 14.8 db became 14.75.  With more resolution in the setting, my dragging of the fader would allow me to drag in smaller increments.

Great, that’s awesome.  Only one thing was getting in my way.  I have tons of auomation and I need to make adjustments.  Regularly I’m dragging lines around and watching my values change in increments of 1.5 db when I want to make half db increments.  WHy oh why can’t I just hold down ctrl like in logic?  I dunno.  

I sat on this for about a week before I had the bright idea of trying what I did for those faders in session view.  In arrange view I increased the view for my channel (finding the horizontal line and dragging it down or up) to resize it.  This worked, and now I could make sub-db adjustments to my automation.

Sweet, finally found it.  But the more I use ableton to mix the more I can’t wait not to mix in it.  There are all kinds of subtle things that get in my way.  I recently discovered that I could add automation lanes for multiple automated parameters (totally sweet).  However, when I adjust things in those lanes, it doesn’t give me a number telling me what I’m changing if I click on a line, even if the line is flat.  So you adjust the line, move your moues curosor to a nearby knot, see the value, and adjust again.  Sure, you’re supposed to adjust everything by ear, but there are plenty of situatoins where just seeing a number while you drag tells you that you went to far.  I set out to make a subtle adjustment, and it turns out it’s huge – just for example.


So, did you read all that?  no fucking way.  Anyways.  The summary:

  • Some days in the studio suck
  • But the task is not impossible
  • There’s a whirl pool of adjustments to make, and they cyclone their way toward a common center
  • Ableton could be a lot better (I could complain about what I don’t like for hours – at least when it comes to mixing, it is so not my favorite tool).