Feeding the Compression Monster

henge stands up, takes his hat off and looks around the room at the discerning eyes staring at him. " I am a compression monster " he states.
Yup I love me some compression. Individual instruments, busses and the 2 buss.
I only have 1 hard comp ( ubk Fatso ) that's used on snare,kick, hat,overheads and bass. Soft comps are working well for everything else since the release of VBC and other soft comps.
So it's little steps of compression along the way on various tracks, comps on the busses and a comp on the 2 buss. Compression monster!
Reduction wise, with the buss comps shaving off a db or two the individual track comps can do much less work. Also have the UBK1 running parallel to the drum buss squashing the shit of the sound! Awesome texture imo.

Hi, my name is Ward and I'm a compress-a-holic.

I went through the phase....where everything got compressed to some degree. Well, maybe not everything--but lots and lots stuff. Now, rather than compression "monster" I prefer compression NINJA. Meaning--I've found the challenge is in finding the LEAST you can do, but in just the right places to make everything balance. I'll give you two tips that will reign in use:

1) never increase the PEAK level with a compressor on a track or buss. Find the loudest section of the track and it should be the same level with the compressor bypassed. If you just insert a compressor....do a bunch of reduction...set the make up gain...you tend to make the signal louder. Which you KNOW sounds "better". So what? You change the fader to compensate, right? What if you decide to question the compressor? You hit bypass and...I can't hear it...wait...it's small...oh, MAN, gotta keep THAT compressor engaged.

2) don't engage any compressors until you've set all the basic levels and panning...done most of your filtering and EQ'ing...and decided what might need saturation--be it a tape, tube, transformer, console--whatever. As you're setting them all, feel free to make mental notes: "I think this acoustic will need some leveling to sit right"...but, don't engage them until the whole mix has been carved and panned an colored. The one exception I make to this is that the lead vocal WILL need compression--so, other than a HPF, I don't do anything to IT until I have it sitting with the dynamics processing it needs. The compressor that's NEEDED will change the frequency balance...maybe add it's own saturation...so, it's in keeping with the same basic minimalist philosophy--the implementation is just opposite.

Here's a couple other points...lots of compressors don't make your mix louder. It makes it smaller, more distorted, and transforms MUSICAL dynamics into electrical dynamics....which tends to necessitate more automation to try to recreate some musical dynamics with MORE electrical dynamics. Speaking to point #2--it's amazing how judicious EQ cuts the need for compression AND makes the compressor more naturally act on frequencies that matter. Take a vocal mic with a HUGE low end...that low end gets SUPER dynamic...super LOUD...makes loud parts of the vocal seem even louder...if you roll some off, suddenly, the vocal that needed 10db of gain reduction needs five to sit the same. Take a snare and hit a tape sim hard--while technically changing the attack and dynamic range, if you try to do the same thing with a limiter, you'll fail. Maybe you'll do something ELSE...and maybe that's cool...but, that's why I put saturation ahead. If a guitar track just has a few icepicky pick attacks and driving it into a console emulation channel just rounds those off--no more need for compression...problem solved with FAR less side effects. No pumping...no pulling down other pieces...no trying to find those frequencies with a notch filter or multiband--and likely making the track duller in the process.

I'm often using dynamics processors now to increase dynamic range (sort of)...for example--a slow attack on a bass guitar will allow the finger/pick attack to retain most of it's musical dynamics, while suppressing or evening the sustain, which is what makes the bass sit or not. If you bypass the compressor, it sounds louder, but less articulate or punchy and even. I use Transient Designer on overheads basically as a simple to set downward expander--allowing the attacks through but suppressing the sustain, which lets me put them higher in the mix without the cymbal washes overwhelming everything. I will generally set a buss compressor to do something similar--and have noted at least 1db of dynamic range INCREASE with it engaged on some tracks...because it's letting some punch through before it engages. But, it's ALSO increasing the RMS and low level details. What's better than increasing the dynamic range AND pushing low level details up?

Anyway...maybe everyone has to go through a "smash everything" phase. I did. I'm here to testify--there's still hope for you.

I don't tend to compress things I don't feel need it.

Once something hits about 5dB of audible gain reduction I find it, outside of effect or utilitarian usage (Ducking for radio programming, side chain compression for EDM), objectionable.

If I'm going for heavy compression, I prefer dirty compressors. Otherwise I'm probably trimming 1-3 dB here and there just to shape things a little. Nothing gets compressed without a specific reason.

Recently I was about to "master" a mix using the WaveBurner program that came with Logic 9. It was for making home brew CD's to give away as a party favor at a New Years Eve gig. Along the way, I'd run a few "alternative" copies in Logic and somehow on one of them, I eliminated two sub busses where I had sent 5 string tracks and on the other, 4 harmony vocals using compression to keep them balanced and to sit in the track. I only noticed I'd removed the compression busses by chance when I was listening, and I noticed the strings had more power, more of a real string sound, Unfortunately, they were still too loud and jumped out dynamically at certain points, overshadowing the vocal. Instead of redoing the bussing to a compressor, ( I was using the UAD LA2 previously), I used automation and lowered all volume approximately .3 db, Cumulatively, this made just enough of a difference for e to finish the mix WITHOUT compression on those 9 tracks, ( there was a pinch of LA2 already printed on the vocal tracks). The mix was simply more alive and retained more interest, and had a more natural dynamic.

So, I'm sort of in the middle. I take Popmann's point of less is more, and appreciate the detailed post he made, as it's opening my mind to more carefully try a few other things to help tame dynamics, like a tape sim or channel strip on one track. I've tried that before with mixed results, but will revisit. But, that said, I still couldn't get as close to what I want without a healthy dose of Slate's VBC, an occasional touch of UAD's Fatso on a guitar track, some more LA2 on vocals, and some A.O.M Limiting when mastering.

I have way less engineering experience than most of the cats here, so I'm still throwing some things on the mix wall, and seeing what sticks.

It just goes to show you that there is no right way to do things when it comes to style. There are definitely right ways to do things like mic placement, gain staging and such, but when it gets down to style, then all bets are off.  It's nothing more than opinion at that point.  Opinion based on one's own prejudices.

Used a Sta Level on bass and vox once and it was awesome. never used their other pieces though...

[/quote]This is awesome.  Just full on hammering the vocals.  I listen to a lot of pop stuff and knew it was hit hard, but this is LEGIT hammered.  

I assume this "stereo stems" that you're talking about here is doubling the drums?  ... 

http://instagram.com/p/hKNqyUFcBU

Anyway, I think it's cool to see someone saying that they hammer the vocals like this when most guys like to talk about tickling the vocals with 3db or something.


[/quote]


Ha. Yep, I'm hammering that thing. You know though, most of the pop stuff I get is already hit pretty hard and often VERY sibilant due to comp settings, EQ, and mic choice. Not from the more experienced producers but the less experiences producers often send tracks that sound, to my ears, like maybe poor mic choice and "it's here so let's use it" EQ and comp technique. Sometimes, like popman says, I'm using the compressor to breathe some apparent dynamics BACK IN! And for me, compression settings like I described are almost ALL ABOUT TONE AND STYLE...very little of it is for transparent dynamics control. I like the urgency that the compressors impart and I DO have to sometimes automate to replace some dynamic movement as the instrumentation changes. I work mostly on heavily stylized, aggressive music. If I was mixing Linda Ronstadt or something I probably be much more gentle.
It's never about just reducing dynamic range or a need to USE the gear I PAID FOR, it's about stylizing the tones and the envelope behavior of the sounds...ya know, getting it to BREATHE.
There's certain envelope behaviors that just don't exist in a transparent capture of a great performance. Sometimes you gotta get it breathing.
Always looking for 100 ways to make something 1% better.
The stems are really so everybody is performing to a more finished sounding drum kit.
But we did triple the damned kit on 1 song...oy....THAT is hard to do well. I'm just gonna have to COMPRESS IT ALL!!!

Monster status?

+1

In my experience, the STA is transparent enough at any reasonable setting for vocals/bass/acoustic guitar. IMHO, the mojo it imparts to the incoming signal is hard to define - everything just sounds better through it.

I gotta say that I use more and more compression as time goes on, but I use it more wisely and as Noah says, I use it for style and color now rather than just leveling out audio.

Flamed?..Why?!...I think we've got a winner!

Sometimes I hate you guys... not only am I trying to find the available funds and justify the expenditure of a STA-Level, now I am trying to also find and justify a 176.

thanks.