The compression question thread

Cool, thanks a ton dude, this is really invaluable stuff! :worship:

I added this part to my above post just recently, so I guess you didn't see it (judging by the fact that it didn't show up when you quoted me), but here it is again:

And you're also saying that the makeup gain has the exact same effect as turning up the track fader when mixing ITB?
 
Though to be honest I don't see how "the compression will bring up the decay of the kick to a more constant level" just because the attack is delayed...as detailed above the makeup gain is still going to be acting on the initial transient.

The attack will come through unaffected if you slow it enough, then the compressor clamps down on the decay. The makeup gain brings whatever is compressed (the decay in this case) back up by the specified amount for the specified time. The release is the compressor letting off the signal by the specified time/amount.

oh wait, i'm not sure if this was a question or not?
 
The makeup gain affects the whole signal, not just the compressed audio, that's why I can't see how the decay can be 'brought up to a more constant level' this way.

Cheers though...
 
The makeup gain affects the whole signal, not just the compressed audio, that's why I can't see how the decay can be 'brought up to a more constant level' this way.

Cheers though...

Yeah, I think that's the active ingredient here, and thus why it strikes me as strange to read about people doing the "set the attack to let the transient through" thing.

So really, when would you not want the shortest attack time guys?
 
When you want to accentuate the attack of the snare/kick/whatever. The key is to have very little makeup gain....you're in fact lowering the volume of the tail/decay etc and causing the peak transient to be louder and stand out more.
 
Well I was more referring to vocals/bass actually, instruments without much in the way of transients (unless we're talking slap bass, but fat chance I'll be recording that any time soon :lol: ), so that's where I thought having the fastest attack would be advantageous; what sayest thou?

You are recording only vowels? Some consonants (atleast in Finnish language) like S,T,D,P,K have that same kind of violent spiky characteristics as for example drumhits compared to vowels which are more like a pile of sausages. So the compressor can remove the rawness from the vocals, and the pick sound from the bass, if the attack is too fast.
 
I don't think those were directed at you, you posted while I was still typing mine.

I don't have the api plugs but from looking at the screenshot I figure: release knob is measured in seconds, and when you turn it fully to the right, you switch over to the other knob which has even shorter times. doesn't really make sense to me, but its copying a classic hardware design

ah yeah i mean i know it has shorter times but it just never made sense to me and the word "variable" there throws me off i keep thinking maybe you pull the knob or something to make it auto release. variable just sounds like auto release to me but i get by without it i do like the auto release feature sometimes.




btw off tangent from my personal question this might help others as it helped me


http://mixonline.com/mag/audio_andy_wallace/


some great tips for mix buss compression. i basically copied his suggested settings and just kind of run with it and adjust as neccesary and use the the -4 db as a guideline. i don't really like to tinker with the master buss compression as much on a real mix as i am not experienced enough but i will toy with it on the drum bus as it sounds phenomenal. like slighty above -4db of GR with auto release and about 10ms attack. the waves ssl comp really shines here. just figured that article might help someone out as well. keep this thread alive! and stickied!
 
A few notes:

Compression, as hinted at and stated earlier, can be used to make decay 'hide' and bring out attack; if you don't like ringy decaying snare you can just set the attack late and have the compressor pull the decay to a lower volume, setting the release time about as long as you can so that you aren't cutting out the next hit, or just gate it if you really want to cut the tail off and worry less. After doing this you'll have a very big 'smack' and a whole lot less tail; compressing with a fast attack and release, limiting, or clipping will then get the attack to behave the way you want it in the mix.

Attack and release are also important tools against pumping - unless you want the volume to sound like it's being jerked around like a porn star's date, you want your attack and release to be set so that your compressor can handle what you're asking it to do. There are a lot of factors in that, of course, but attack and release are the first things I go to.

For smashing things quickly, you of course have the option of clipping (an effect that doesn't have to sound scratchy or buzzy; the cranked 'tube clean' on old Fenders at high volumes are an example of this - no 'distortion' is audible but the harmonic content is changed and the sound is noticeably compressed), but for sticking to straight compression you might also want to try lookahead compression units - they 'look ahead' at the signal (hence the name), take a certain average measurement, compare that to other dynamic measurements, and then fiddle with the gain accordingly.

The Slipperman FAQ had an analogy relating well to this, worth looking up to be certain. To borrow a few of the Smurfs that were taking care of our loudness in the Getting Your Loudness thread ( http://www.ultimatemetal.com/forum/andy-sneap/278894-getting-your-loudness.html ), I'll abuse the train of thought slightly:

You have Smurfs in a box with knobs. Their sole purpose in existence is to fiddle with these knobs - knobs that happen to control volume, of all things - according to your wishes. You've given them the Threshold setting, which tells them when to fiddle with knobs; you've given them the Ratio setting, which tells them how much fiddling to do. However, I don't think of Attack and Release knobs as parameters they're given directly; the Smurfs actually have intravenous tubes that sustain them, because feeding compressors would make your clients think you're crazy, and into these intravenous tubes we shoot a dose of amphetamines corresponding to the Attack and Release settings.

In this box, one Smurf just sits there and waits for the threshold to be hit, and when it gets there he smacks the fader down according to the Threshold, Ratio, and Attack parameters; another waits for the signal to drop below the threshold, and smacks the fader out of the first Smurf's hands and back to unity. When you shove more liquid energy into their veins, they move faster... and start to get a little fidgety. And then a little more fidgety. And then they start getting into speed-fueled rages about where the volume should be, fighting over the knobs ruthlessly and maniacally, until they get worn out and lose track of what they're doing while arguing over imaginary Smurfettes and eventually trying to have their way with cute blue hallucinations in fuck-me boots waiting to be abused on the volume knobs... which sounds like pumping to you.

So be nice to your Smurfs - they get lonely in those boxes, and when they start imagining females things are bound to get ugly... giving you a whole new mental image of what's going on in your compressors when things start pumping.

Jeff
 
Jeff - that may be your greatest post (and analogy) yet...many thanks as always for the laugh(s) and info! :wave:
 
Most compressors have a "Gain reduction" meter, which basically just tells you how much dBs you're shaving off in TOTAL, so all parts (attack, release, ratio and threshold) are taken into account in this.