this is the thread where you tell me how Compression works and how to get the best out of it

http://www.digido.com/

Okay. Everyone's seen pictures of sound represented as a wave-form; if you look at one, you can see that the wave's amplitude is constantly changing. What that means to a listener is the sound's _level_ from instant to instant. If you raise the level of something, it gets louder; if you lower the level, it gets quieter.

When you record a voice or an instrument, it's not going to maintain the same level, even for a fraction of a second. There are fluctuations in level going on all the time, and some of them can be intense and distracting. In most cases, you're going to want your recording to sound more consistent than its source actually is--or at least to make sure that your recording's major inconsistencies in level or in sound are intentional.

A compressor is essentially a tool to make levels more consistent by limiting their range. To use compression, you first decide where you want it to start working. You pick a threshold level--say, 12 deciBels (dB)--and set the compressor to affect everything above that level (and ignore everything below it). Then you set the ratio for the compressor--3 to 1, say. So when the compressor processes the recording, for every sound that's louder than the threshold, it will take the _amount_ that its level is higher than the threshold, and reduce it according to the ratio. If a compressor is set as we've described it here, and a particular peak of a recording is 18 dB, the compressor will change that peak so it becomes 14 dB: (18 dB minus 12 dB) times 1/3, plus 12 dB.

So compressors make things quieter, right? Technically, yes, but in practice it's the other way around. The reason you use compression is so that you can then raise the _overall_ level of the recording so that the loudest parts are as loud as they were before. What you've effectively done is making the quiet parts louder--bringing the stage up, so to speak, instead of bringing the ceiling down.

That's a really, really useful trick to be able to pull off. Compression is like salt: a little of it makes everything sound better. Compressed voices sound more authoritative; compressed instruments sound more precise and energetic. Done properly, it gives sound more oomph. Analog recording media actually have a certain amount of compression built into the format. Recording engineers _love_ compression--very often, they'll apply a little bit to most or all of the instruments in a mix, and then a little more to the whole thing. But compression also affects the sound--changing the waveform of a sound, no matter how slightly, alters its timbre, and there's no getting around that. If you apply compression to certain kinds of sounds on the low end of the spectrum, especially bass guitar and some drums, you can actually hear it making a whooshing sound as it kicks in. That's a bad thing unless it's part of your artistic intention, which it very occasionally is.

To get it to sound good, you need to play with the parameters depending on what you want to do. The threshold is the level at which the compression should happen. (ie how loud until you reach for the volume). Ratio refers to the ratio of output to input, or how many times quieter to make the output when the input passes the threshold.

Attack is the time between the signal passing the threshold and the compressor pumping it down. Release is how long to keep the volume down. To get a feel for it, get a piece of sound thats quiet and then suddenly gets louder. Change the attack/release until you can hear their effects.

Dunno if that helps at all. Hopefully some other ppl will add some more...

i already regret asking this next question because i know the answer will be just play around and see what sounds best,,,,, but is there a sorta standard setting to make stuff sound the best.

how about a way to make it so that you get noticable differences in sound (like the way Douglas was saying you get a wooshing sound)?

i'm sure it all depends on the incoming source, but i just haven't heard any difference in the sound and i've done a bunch of knob twidling

thanks again.

I like it because it makes my voice sound like the "This segment of the Late, Late Show with Craig Kilbourn is sponsered by...." announcer.

http://www.soundonsound.com/sos/1996_articles/apr96/compression.html?session=07e8002f8f04f23a5d63391c0119d701

(and if the link fails, go to www.soundonsound.com and search for "compression".) -sound on sound is an amazing resource, for me the best thing on the internet, and the best engineering mag in print too.

SOURCE
ATTACK
RELEASE
RATIO
HARD/SOFT
GAIN RED

Vocal
Fast
0.5s/Auto
2:1 - 8:1
Soft
3 - 8dB

Rock vocal
Fast
0.3s
4:1 - 10:1
Hard
5 - 15dB

Acc guitar
5 - 10ms
0.5s/Auto
5 - 10:1
Soft/Hard
5 - 12dB

Elec guitar
2 - 5ms
0.5s/Auto
8:1
Hard
5 - 15dB

Kick and snare
1 - 5ms
0.2s/Auto
5 - 10:1
Hard
5 - 15 dB

Bass
2 - 10ms
0.5s/Auto
4 - 12:1
Hard
5 - 15dB

Brass
1 - 5ms
0.3s/Auto
6 - 15:1
Hard
8 - 15dB

Mixes
Fast
0.4s/Auto
2 - 6:1
Soft
2 - 10dB (Stereo Link On)

General
Fast
0.5s/Auto
5:1
Soft
10dB

the 2nd verse of the 2nd song on http://image.allmusic.com/00/amg/cov200/drf400/f474/f47447zhvt1.jpg everything drops out (or way down) except the guitar (buzzing, kinda like a mellow 1984 bob mould tone) and vox which are over-compressed to the point of trebly distortion... but then out of all of that, a really soft harmonics that sound almost like bell-y keyboard tones emerge in the back half of the verse. i always liked that part. i need to listen to this again.

hands on real knobs through real circuitry will give you an
experience that you can extend to things like reason's little
comp's.

or if that's too much work another alternative is the 1176 VST
plug in (bomb factory):
with attack at 0 it's just ridiculous. you'd have to be near
dull to not notice this on a snare.

http://www.fmraudio.com/overview.html