Compression help

[daft cunt]

I've read the basics of compression: what's supposed to do what but still I can't figure why the threshold has to be set at x dB for a bass and y for a snare. Same for the ratio...
So if someone is up for a little explanation or possibly a tut that'd be great.

[komega]

yes please, thx

[Sharmaji]

threshold is where on the volume scale the compressor starts working. if you figure that 0db is the loudest you can get, you slide the threshold down and that's where it starts compressing material. anything under the threshold goes through unaffected; anything over gets compressed...

at...

the ratio you set it to. so say you've got your threshold set to -10... something comes through that's like -6.. 4db over the threshold, right? it's gonna get compressed. if your compressor is set to a 4:1 ratio then it'll get compressed down to 1db.

so basically a compressor turns things down by a certain ratio... only thing that's confusing is that some compressor plug-ins, like the normal one in reason and the silver comp in logic, automatically make things louder by the same amount... so if you knock off 3db from a signal, it turns it up by 3db... which is weird.

think of it this way: the threshold is when the compressor shows up for work; the ratio is how hard it works. attack is how long it takes to ease into the day, release is how long it lingers around.

[djdinyr]

Compression is an art that takes practice to do well. Here are some starting points that I use:

Threshold: for drum hits (kick and snare) and bass, I usually set it somewhere between -10 db and -15 db. You should be able to find a good threshold in that range. If you go higher, the compressor will probably not kick in. If you go lower, too much compression will result and the sound will get flattened. However, this all depends on what you're compressing.

Ratio: again, for drum hits and bass, I usually set it between 4:1 and 6:1. Kicks usually get the highest ratio, and the bass the lowest. If you go lower, the compression will probably be too weak (2:1 ratio is good applied to the song as a whole versus an individual track). If you go higher, the compression will be too much (more like a limiter).

[komega]

thx boys

[jahtao]

Take all advice with a pinch of salt. First hand experience of compression is a must. Experience over time comes next. For understanding compression its all about THE GRAPH. Let me see if i can find one for ya......

[jahtao]

Here's something to keep you going but its not 'the one'


http://www.petethomas.co.uk/recording-compression.html

[jahtao]

Right, having done nothing more than cast my eye over this for 2 seconds, i declare it to be the ultimate guide to compression... and yes i know its 7 years old

http://www.soundonsound.com/sos/dec00/a ... ession.htm

remember compression is an important part of getting a pro sound but it is easily over used, used wrong and most importantly, just coz you want your shit to sound pro doesn't mean it needs loads of compression or even needs it at all, its ok if you like it with no compression.

[decklyn]

Essentially compression changes the ratio between the peaks and averages of a signal. This is what the ratio setting is on the compressor.

Now it doesn't just constantly do this, tho. There are a few other parameters on there too. Threshhold is the level at which a signal must be in order for compression to be applied.

Attack is the time in which it takes for the compressor to actually kick in. So if at 1 second the signal goes above the threshhold and you have a 30ms attack, the compressor will not kick in until 1 second 30 ms.

Release is the time which it takes for the compressor to release after the signal drops below the threshhold.

Hard knee and soft knee describe how the compressor kicks in - iether instantly or slowly. (double check this - this is just what I understand).

Don't really know what else to say.

Hard limiting is where you set the compressor to an infinite to one ratio, so that as soon as the signal goes above the threshhold it's limited to that threshhold level.

[subframe]

^^^ that's correct, hard vs. soft knee describes the way the compressor goes from not-compressing to compressing. It can be either a hard corner exactly at the threshold, or a softer transition from uncompressed to compressed. Each sounds better sometimes.

Jahtao had it right - there are no hard and fast rules, take all advice with a handful of salt.

Something I found helpful was to whack up the ratio and lower the threshold to extreme settings. Then play with attack and release on various signals. Drums, basses, etc. Then back off the ratio and threshold, see how that changes what happens. Then do the opposite. Set attach and release to very fast settings, and vary the threshold and ratio. Then do it all at once

[decklyn]

^^^ that's correct, hard vs. soft knee describes the way the compressor goes from not-compressing to compressing. It can be either a hard corner exactly at the threshold, or a softer transition from uncompressed to compressed. Each sounds better sometimes.

Jahtao had it right - there are no hard and fast rules, take all advice with a handful of salt.

Something I found helpful was to whack up the ratio and lower the threshold to extreme settings. Then play with attack and release on various signals. Drums, basses, etc. Then back off the ratio and threshold, see how that changes what happens. Then do the opposite. Set attach and release to very fast settings, and vary the threshold and ratio. Then do it all at once

Thanks sub frame. good advice will try.

[daft cunt]

Thx very much everybody

[deadly_habit]

all about sidechained compression

[jahtao]

all about sidechained compression

blap

[daft cunt]

all about sidechained compression

1° What's sidechain?
2° Should the compressor be put at the output of the generator/sampler or at the end of the effects chain?

[demento]

check sos or google sidechained compression or sidechaining

[decklyn]

1. Sidechain compression is where you use sidechain inputs on a compressor from one signal to compress another signal. Usually the sidechain inputs have eq or lp/hp filters. This can be used for a few different things. probably the most common usage is a technique called "ducking."

Ducking is where you use the kick in a track as the sidechain input to compress the bass.

Essentially what happens is what when the kick hits, the bass gets turned down, so that any muddying of bass frequencies is avoided, and the kick punches through the mix. It's used more in music like house where the bass and kick frequencies are both very low end. Drum and bass sidechaining is used quite less frequenty, as the bass frequencies tend to be below that of the kick frequencies.

There are other uses. In house you often hear the kick routed to sidechain inputs to compress pads in order to give them the track that boomy stadium feel.

Another use is in using delay on vocals. If you just apply heavy delay, sometimes things can sound a bit cluttered on vocals. So if you want heavy delay, but clarity while the vocalist is actually singing the vocals, you can apply compression to the delay signal using the original dry sample routed to the sidechain inputs so that the delay signal is turned down whenever the vocalist is singing, but then there is heavy delay carying on when the vocalist is not singing.

2. Usually you would compress the dry signal after EQing and then apply effects after this. Compression is almost aways used as an insert, so any send effects would be affecting the compressed signal anyways.

3. Also learn about parralel compression.

[daft cunt]

1. Sidechain compression is where you use sidechain inputs on a compressor from one signal to compress another signal. Usually the sidechain inputs have eq or lp/hp filters. This can be used for a few different things. probably the most common usage is a technique called "ducking."

Ducking is where you use the kick in a track as the sidechain input to compress the bass.

Essentially what happens is what when the kick hits, the bass gets turned down, so that any muddying of bass frequencies is avoided, and the kick punches through the mix. It's used more in music like house where the bass and kick frequencies are both very low end. Drum and bass sidechaining is used quite less frequenty, as the bass frequencies tend to be below that of the kick frequencies.

There are other uses. In house you often hear the kick routed to sidechain inputs to compress pads in order to give them the track that boomy stadium feel.

Another use is in using delay on vocals. If you just apply heavy delay, sometimes things can sound a bit cluttered on vocals. So if you want heavy delay, but clarity while the vocalist is actually singing the vocals, you can apply compression to the delay signal using the original dry sample routed to the sidechain inputs so that the delay signal is turned down whenever the vocalist is singing, but then there is heavy delay carying on when the vocalist is not singing.

2. Usually you would compress the dry signal after EQing and then apply effects after this. Compression is almost aways used as an insert, so any send effects would be affecting the compressed signal anyways.

3. Also learn about parralel compression.

Thank you for that!

[decklyn]

No problems. If you understand that, then I'm sure it's evident, but I think I forgot to mention the uses of LP/HP or EQ on sidechain inputs

If you have a percussion buss and you only want the bass frequencies of that buss to compress the bass, then you would EQ or filter the signal accordingly.