I think I just had a compression revelation.

thelowerlip

thelowerlip

New member
I often use a small amount of compression to level out dynamics.
I was just watching the vst compressor that came on my DAW and I noticed that when a thick chord was played on a piezo, the compressor was all over it... but not so much on the high, snappy notes.

Well of course!!!
The low frequencies will reach the threshhold before anything else does.
So... if you're trying to level out some high frequency pops, you need to eq before compression or use a multiband.:facepalm:
 
The "normal" way would be to HPF the input to the sidechain insert.

Likely, you'd be EQ'ing before the compressor anyway in most cases. Still, HP'ing the sidechain insert to (whatever -- I tend to be around 150Hz on most things) is going to be an awfully big help...
 
thelowerlip said:
Well of course!!! The low frequencies will reach the threshhold before anything else does. So... if you're trying to level out some high frequency pops, you need to eq before compression or use a multiband.
Click to expand...

Exactly. Often two EQs are needed: One to filter out mud and rumble and other stuff you know is not wanted, then you compress, then you can add another EQ to actually shape the sound. If you EQ for tone before compressing, the compressor tends to undo the EQ.

--Ethan
 
or, go to the root of the problem, and address your tracking.

down on the gain structuring....
more select Eq'ing.....

and smarter compression.


youi want to address the problem GOING IN...

not coming out...
 
thelowerlip said:
The low frequencies will reach the threshhold before anything else does.
Click to expand...

That depends entirely on the source material. What reaches the threshold first is whatever is loudest.

But if you're going to compress you might as well compress the frequencies you plan to keep rather than let the compressor react to stuff you're going to filter out.
 
Interesting topic.

If you're really just trying to clamp down on the hi-freq dynamics, I think the multiband compressor is definitely the way to go. For vocal sibilance, I usually severely compress/limit above around 4kHz with a very fast release. And the threshold setting is very important - during parts where I don't hear any sibilance, I make sure the compressor in that band is not doing anything, but when sibilance comes in - the gain reduction meter would light up all the way off the meters. I don't know why, but it makes me happy seeing the gain reduction meter dance along with the sibilance. =D

Anyway, using a HPF before compressor to prevent bass-frequencies hogging the comp is a definite must-do in my ears. I typically HPF every track in the mix, each at a different frequency depending on the content. And then I'd lightly compress almost every track, before I do any EQ for tone-shaping. I don't know if it's a good practice, but I find this a great way to remove mud and excessive dynamics.

To the experts, may I just ask:
1) what kind of slope you guys set your HPF to? Gentle slope or do you make the bass frequencies fall off an abrupt cliff?
2) some people actually advocate setting a low threshold and low ratio on the compressor, such that the track is being compressed all the time - lightly at soft parts, and more at loud parts. Does this actually help alleviate the problem of "bass frequencies hogging the compressor" at all? Or is this even a good practice?

Cheers!
 
The steeper you set your filter slopes, the more the resultant audio will be prone to distortion.

Like an awful lot of things we do, it's a trade off between achieving an acceptable sound coupled with minimum side effects.
 
1) what kind of slope you guys set your HPF to? Gentle slope or do you make the bass frequencies fall off an abrupt cliff?
Click to expand...
It depends...
2) some people actually advocate setting a low threshold and low ratio on the compressor, such that the track is being compressed all the time - lightly at soft parts, and more at loud parts. Does this actually help alleviate the problem of "bass frequencies hogging the compressor" at all? Or is this even a good practice?
Click to expand...
It depends...

Seriously though -- It depends. If you're trying to "bring the lows down gently" then maybe 6dB/octave will do. If there are problematic anomalies maybe an elliptic filter at 100dB/octave is the only thing that'll cover it. It depends.

Same thing with compression. Either way, the sound tells you what to do. If it's not telling you what to do, you aren't listening.
 
Ethan Winer said:
Exactly. Often two EQs are needed: One to filter out mud and other stuff you know is not wanted, then you compress, then you can add another EQ to actually shape the sound. If you EQ for tone before compressing, the compressor tends to undo the EQ.
--Ethan
Click to expand...

But sometimes having the compressor undo your equalizer is exactly what you want. You can take a kick and boost low end like a champ so the compressor clamps on it, transforming the kick into a round, big bodied pillow which envelops the ears with heavenly bliss. Equalizing after the the compressor may leave you with an overly transient low octave that will eat up your headroom and pump the master buss compressor if there is one.
 
japancakes,

I can't see how your suggestion works. If you had wanted more compression on the kick, you could simply reduce the threshold?
 
You must log in or register to reply here.
Back
Top