Compression Experts Please Come Forward

  • Thread starter Thread starter tkingen
  • Start date Start date
tkingen

tkingen

Djembes Rock
For the newbie (and not so newbie) compression can be a confusing thing. Even after reading explanations of how compression works it can still be difficult for a novice to put the knowledge to practical use.
So, to all of you who are well versed in using compression - what goes through your mind while you are deciding if you need to compress and how do you decide how to set threshold, ratio, attack and release?
What I'm looking for is a description of your inner dialog that helps you achieve your goals.

Thanks
 
really it comes down to just being able to hear it. and that comes with the experience of using it over and over.

I remember when I had people telling me that they could hear my compresser pumping in my mixes, and I just couldn't hear it. I couldn't tell if it was my ears or my speakers or both. I realized later it was both.
What you should do is get some good monitors (no PC speakers won't cut it) and just sit in a quiet room and listen to a short vocal section looped over and over. First, just close your eyes and leave the compression off....sit and listen to the subtleties of the vocal, their dynamics, the way they sing the phrase, etc. Don't listen to pitch content, the words, or anything...ignore all that and listen to the dynamics. After awhile you'll be able to hear and almost feel the inconsistencies of their voice. This will be your reference track.

now this is where you'll get differing opions on the topic...because it all depends on the sound you want to get out of it.
But after listening, I like to figure out where the threshold should be. I pick a spot that sounds the worst/most inconsistent and watch the gain reduction meter so that the unattractive jumps in volume are kept down to where the good sounding dynamics are at. You can add more or less to taste depending on how dynamic the vocalist is and how much you want to squash them. Ratio I like to start next around 3:1 to 4:1...a nice middle number...and go from there. The higher you go, the less dynamic, the lower, the more subtle the compression. You get around 8:1 and you're starting to turn the compressor into a limiter. At this point, you'll start to hear the vocal getting squashed so I'll turn up the make up gain a little bit to get the overall volume back to where it was when it came in...then adjust attack and release.

Attack and release are another one of those things that you just have to listen and feel. I set them at extremes to start off at (like 1.5s attack) and pull it back until it's almost impossible to hear the compressor kicking in. If you set it to an extreme first, you should be able to hear what you DON'T want and then it'll be easier to find what you DO want. Same thing with release. A lot of people like to keep a general rule of thumb being "faster song, faster attack..slower song, slower attack".

Anyway, hope this helps some. Just my way of doing things. But I'm not by far an expert at all. :cool:
 
Last edited:
bennychico11 said:
really it comes down to just being able to hear it. and that comes with the experience of using it over and over.

I remember when I had people telling me that they could hear my compresser pumping in my mixes, and I just couldn't hear it. I couldn't tell if it was my ears or my speakers or both. I realized later it was both.
What you should do is get some good monitors (no PC speakers won't cut it) and just sit in a quiet room and listen to a short vocal section looped over and over. First, just close your eyes and leave the compression off....sit and listen to the subtleties of the vocal, their dynamics, the way they sing the phrase, etc. Don't listen to pitch content, the words, or anything...ignore all that and listen to the dynamics. After awhile you'll be able to hear and almost feel the inconsistencies of their voice. This will be your reference track.

now this is where you'll get differing opions on the topic...because it all depends on the sound you want to get out of it.
But after listening, I like to figure out where the threshold should be. I pick a spot that sounds the worst/most inconsistent and watch the gain reduction meter so that the unattractive jumps in volume are kept down to where the good sounding dynamics are at. You can add more or less to taste depending on how dynamic the vocalist is and how much you want to squash them. Ratio I like to start next around 3:1 to 4:1...a nice middle number...and go from there. The higher you go, the less dynamic, the lower, the more subtle the compression. You get around 8:1 and you're starting to turn the compressor into a limiter. At this point, you'll start to hear the vocal getting squashed so I'll turn up the make up gain a little bit to get the overall volume back to where it was when it came in...then adjust attack and release.

Attack and release are another one of those things that you just have to listen and feel. I set them at extremes to start off at (like 1.5s attack) and pull it back until it's almost impossible to hear the compressor kicking in. Same thing with release. A lot of people like to keep a general rule of thumb being "faster song, faster attack..slower song, slower attack".

Anyway, hope this helps some. Just my way of doing things. But I'm not by far an expert at all. :cool:
Click to expand...

Good stuff, one of the best thing you can do for practice is over compress and make it pump and squash it, than play with the timing and threshold, do this with different instruments and vocals, It's like taking a new sports car out for a spin, unless you test it in a bunch of different situations you don't know what it can and can't do, and don't get caught up in the "numbers", sometimes I look at what my compresor is set at and think "that can't be right" but hey it sound good, some tracks will need a "limiter" like ratio but the threshold set to only cut off the "peaks" others will need a lower than "normal" ratio and more agressive threshold to smooth "things" out, some times you will want to hear the compresion and use it as another instruments, Practice practice pratice...
 
Last edited:
Lemme pull this in a different direction -

Compression takes practice, yes. Compression ALSO takes and understanding of what you want.

Throwing a chorus or a flanger on something is one thing - You can throw it on and twirl the knobs for a bit until you get what you want - After all, you want some chorus or flanging.

Compression is a different story - With dynamics, you need to have a clear "end game" in mind. You need to listen to a track and think how you want it to sound. So first, you have to think "hey, this needs compression." more than that, a specific type of compression - Soft, hard, crushing, limiting, with the groove, against the groove... Just throwing a compressor on and fiddling with the knobs until it sounds right is *normally* a road to ruin. Without at least a basic "automatic" thought of some rough settings for a specific effect on a specific sound, you'll be compressing yourself into a hole in no time.

So with that in mind, it goes back to practice - Practice on things you're NOT working on. A really dynamic drum loop is nice - especially if it has a lot of room sound. Play with the threshold and extreme ratios. (-40 @ 1.1:1) is going to have an entirely different effect than (-5 @ 10:1). Same with long or short attack and release settings. They can either help or seriously hinder the sound you're looking for. Keep a close ear on how the controls interact with each other and the effect they have on the track.

After you understand the concepts and controls in that way, your "inner dialogue" as you put it, will simply tell you "That needs a few dB of really soft gain reduction" or "this need to be limited 3 or 4 dB off the top."
 
well you just got some really good advice.

personally (I know NOTHING) I compress when I hear something jumping around volume-wise and I want it to stay put. what that "something" is is going to determine the settings for me to get it to stay put. also, it might depend on what part of the "something" you want to stay put. so, say you have a bass thats all over the place. it wont "sit" in the mix because some notes are loud and some are barely audible. you also have this higher string sound and a lower boom or woofer sound, and when you bring up the bass to where the string sound sounds good the boom is just out of control. so you put some limiting and bring it down to where the boom is under control and solid, and the string sound is able to come out. or maybe you want the boom gone outright. so you set a slow attack which lets the string sound through, but then clamps down 4:1 and stays down with a slow release immediately after that attack. the boomy sound is now brought down in volume based on the time of the sound, whereas with the limiting example you brought the volume 'ceiling' down so that a less audible sound is heard more clearly. these are just two ways to attack a problem with compression. some might say 'you could have used eq' or that my settings dont make any sense but it all comes down to what you want it to sound like and you only have an "idea for what the settings should be" after fudging with it in enough different circumstances. i hope this helps.
 
Also (for the DAW) don't be afraid of geting in there with the editing tools and pick an offending word or phrase and manualy set the envelope or fade to get it under control, that way you won't have to overuse compression on the track just so that word/phrase does'nt "stick out", there's also nothing saying your limited to one compressor on a track.
 
A refreshing informative discussion about individules and there discoveries about compression and how they managed to make it work for them.

Whoa......lovely and not (at least to this point) one word about needing to up grade to a distressor or an LA2A or an Avalon.

I feel the tide turning.....
 
Joseph Hanna said:
A refreshing informative discussion about individules and there discoveries about compression and how they managed to make it work for them.

Whoa......lovely and not (at least to this point) one word about needing to up grade to a distressor or an LA2A or an Avalon.

I feel the tide turning.....
Click to expand...

Actually, I think the FMR RNC is an excellent piece to learn on.
 
Anybody care to chime in on chaining compressors? I often use a compressor set with a lowish ratio, to get 3 or 4 dbs reduction, then take that and hit it a little harder (higher ratio), with a little less reduction.

I find it makes the compression less "obvious".

Of course, this is on Jack Endino's list of ways to over-produce a record!

How To Over Produce a Rock Record
 
tkingen said:
For the newbie (and not so newbie) compression can be a confusing thing. Even after reading explanations of how compression works it can still be difficult for a novice to put the knowledge to practical use.
So, to all of you who are well versed in using compression - what goes through your mind while you are deciding if you need to compress and how do you decide how to set threshold, ratio, attack and release?
What I'm looking for is a description of your inner dialog that helps you achieve your goals.
Click to expand...


I think the one thing that pretty much broke it all right through for me is the understanding that sound has density.

Not only does sound have density, but it also has shape. And when I say sound, I guess I'm refering the dynamics of sound as having coordinates in time; there is a beginning to a sound, a middle and an end. And along this continuum, the volume will vary.

Attack, release, ratio, etc. all have to do with what part or parts of the sound you wish to emphasize. In a vocal track, for instance, a slower attack time will have the effect of emphasizing the beginnings of words. This is because the initial consonant sounds the mouth makes are allowed to "pass" before the compressor kicks in ... thus, raising the volume of these beginnings of words in relation to the rest of the word. Ditto for the attack of the pick, or finger, on a guitar string.

Similarly, you can emphasize the endings of the words by using a very short release time. The release time, as you probably know, is the amount of time it takes to bring the volume back up after compressing. If you bring it back up fast enough (by using a shorter release time), it can raise the volume back up before the ends of the words are formed -- thus raising their volume.

By combining the two -- slow attack, fast release -- you enhance the anunciation and clarity of the vocal track, since the beginnings and endings of words are louder.

By doing just the opposite (short attack, long release), you can emphasize the "meat" of the note.

This kinda' stuff is really important for drums, because it's such a dynamic instrument. Take a snare drum. Your stick hits the snare, making a really short, sharp "snap!" sound due to the impact of the stick on the drum head. What follows that is going to be the resonance of the snare's body, followed by the sound of the room reflections. Under normal conditions, the initial snap of the stick on snare might be so much louder than the ensuing resonance and ambience, that you lose a lot of that extra sound. Especially in a busy mix. All you might hear is a wimpy "ting!" or "tss!"

By using a short attack time, you can use a compressor to really bring out the rest of that sound that follows the stick hit. By doing this, you actually make the sound denser. You've literally compressed it so the loud parts and soft parts are closer in volume. And when a sound is denser, the human ear perceives it as being louder. This has to do with volume and our perception of time. Using the snare sample again ... by drawing out the resonance of the snare, followed by the room reflections, you've elongated the snare hit.

You can see this stuff, literally, by looking at the wave form of a drum hit in your audio editor, pre and post compression. A snare hit who's initial impact jumps to "0" and instantly dies out won't sound as loud as a snare who's impact goes up to -5, but has some sustain by several milliseconds of resonance.

And therein lies the danger. Because of this, it's very easy to fall in to the trap of overdoing it. By overcompressing, you basically kill any differences between the louder and quieter parts of a sound. Thus destroying any liveliness.

Well-compressed: POP!pppppPOP!ppppBANGboomBAM!

Squashed to hell / over-compressed: POPPPPPPOPPPPPBANGBOOMBAM

In the squashed version, you just lost the points of emphasis and exclamation, along with any emotion that was contained within it.
 
... and that's why people end up saying something like 'compression kills everything'...

Really good tips above -- listen to them.

Two more little notes:
If you don't have experience, you'll not be able to set the compressor (maybe some presets will work with slight modifications).
If you don't understand compression, you'll never be able to understand when to compress a single track, when to compress a bus/group, how to sidechain,...

aXel
 
Then there is thee nubs approach to compression. Generally I try not to use it or use it in a subtle and un-obtrusive manner. I listen to a track relative to others. I ask myself is the track relative to the others getting buried? If the answer is yes, I raise the fader until I can hear the buried parts at the level I want to. Now I ask myself are the loud parts too loud? If the answer is yes then compression is in order. In most applications I go for the fastest attack I can get, with a release varying on the nature of the sound for maximum neutrality, i.e. pegging down peaks a bit then getting out of the way as fast as possible. Lots of sustain slow release, not much sustain faster release. I am looking for neutrality, i.e. the track sounds "clear." The ratio varies with the peak/valley ratio. This is my general approach, but of course compression can be used for effect: Severe squashing for bowel wrenching kick, bringing up room sound, bringing up pick attack, increasing vocal intelligibility by emphasizing consonants, increasing sustain (or vowels in vocals), increasing vocal breathiness, etc. Also compression is a reason for recording tracks with low noise floors. I know you wet behind the ears newbies record something and you think to yourself "aaaaw this is quiet enough" But lo and behold, get some compression all up in dat sweet thang and all of sudden those quiet rustles ain't so quiet. Sucks don't it!
 
I haven't done it for some time but this thread is getting the cut and paste into a word document treatment.

Some really excellent stuff, thanks folks!
 
tkingen said:
For the newbie (and not so newbie) compression can be a confusing thing. Even after reading explanations of how compression works it can still be difficult for a novice to put the knowledge to practical use.
So, to all of you who are well versed in using compression - what goes through your mind while you are deciding if you need to compress and how do you decide how to set threshold, ratio, attack and release?
What I'm looking for is a description of your inner dialog that helps you achieve your goals.

Thanks
Click to expand...

First thing to figure out is whether you want to compress to control the dynamic range, or if you want to compress to get an effect.

If i need to control the dynamics of something, say a kick drum.
Ill start with everything to zero.
Zero thresh.
Fastest attack.
Slowest release.
1:1 ratio.
Thats what id consider "Zero".

From there id bring in the lowest compression ratio.
Then start to pull down the threshold till i hear it working.

Then ill adjust the attack time till i like what im hearing.
If i feel that the compression isnt strong enough, ill up the ratio a bit more.
Once i feel im whacking the signal correctly, ill begin to mess with the release time.
Slowly moving it towards a shorter release time until i like what i hear.
From there its just fine adjustments, with the afore mention controls to get just what i set out for initially.

As for using a compressor as an effect?

Thats anyones guess. Just play witht the knobs and you might find something cool, or maybe not. Ciompression as an "effect" is more about experimentation than science IMO.

Best of luck,

-Finster
 
Kevin DeSchwazi said:
I haven't done it for some time but this thread is getting the cut and paste into a word document treatment.

Some really excellent stuff, thanks folks!
Click to expand...
Yep. It's nice to see straight forward, well reasoned and well written answers to an important question here. We could use more of that on this site. Thanks.
 
run an L2 on every chanell at about -30, put the output at about -5db
and make sure only to eq after the L2.

makes a great meal. ;)











seriously, i've seen people that think the L2 is the answer to everything.
 
A.R.R.T.

Attack, Release, Ratio, Treshold.

Set it up so it's compressing constantly, with the fastest attack, fastest release, highest ratio and the threshold where it's always working, but working harder on louder parts.

Sounds horrible and tiny yes ? Maybe a little distorted ?

Play with the attack time until you can hear the compressor attacking the start of the sound. If you can't hear it you need better speakers.

When you have the start of the envelope the way you want it then move on to the release...sometimes it's good to audition this with the mix, a bit of rhythic 'pumping' can be cool if it breathes with the track.

Then set the ratio to a point where you don't lose the effect you are going for, but it becpmes less insane..

Next adjust the threshold so it only works on the dynamic peaks you need it to.

NAthan
 
this is some great advice here. comps take the longest to get used to out of any outboard imo. every come has a differant sound and thus will give you a differant end. some a good for smoothing out dynamic edges, some are good for crushing. the only way to get used to it has been suggested here already. i got most used to comp sounds through error, many mixes ended up with pumping hi hat or crashes due to heavy comping of snare and room mics.
unless your using a comp as an effect it shouldnt be overly noticable. slight settings to smooth vox, quick heavy settings to give kick and snare definition.

if you can really here it working youve probaly over done it. it really takes alot of time to feel the comp going as opposed to hearing it. try an rnc, a dbx 160-166 or an art pro vla ( damn smooth sounding soft knee comp) to get a feel for compression.
 
Wow, a heartfelt thanks to all who contributed to this thread. This is the best article on compression that I've seen. The theory makes musch more sense now.
Thanks!!!!!!

Terry
 
You must log in or register to reply here.
Back
Top