Compression VST Question

Started by KrazyKatz, April 24, 2006, 12:48:05

Previous topic - Next topic


Howdy boys.

What does a compression VST do?

I mainly compose classical music so I dont need to flange and "weirderise" instruments and find that the only VSTs I need are graphic equalizers and Reverb. Am I missing something? Im more of a composer than a sound engineer.

Sonic Brilliance Studios


compression is sort of like an automatic volume control which you can set to make loud bits quieter or quiet bits louder (although i think this is called an expander too, lots of plugs do both).
Compression is extremely useful for making things louder on average as it can squash those big peaks you can get and bring the rest of the track up in volume.
imagine a large audio track with one massive peak where, for example, a coincidence makes a certain frequency jump to 0dB but the rest of the track is at half volume. naturally you can single out that peak, reduce its volume and then re-normalise everything to make it louder but what if there are dozens of peaks and you don't have time to remove them all properly? thats where a compressor can do the job for you and squash those parts (and allow you to bring up the rest to a reasonable volume). obviously in doing this you lose a little bit of dynamics but there are almost always a few freak peaks that can be brought down by a few dB without losing anything from the music. It is definitely to be used carefully and sparingly in classical music. pop records  are ultra-compressed /limited (a limiter is like a compressor but you can tell it to never let sounds get above a certain volume) to make them sound very loud all the way through and thus sound better on the radio, doing this to delicate classical material is noticeable and frowned upon! however, with any material, especially classical, containing lots of instruments there will be points where the signal goes much louder than the rest of the piece due to harmonics adding up and where the frequencies coincide exactly you get a large jump in volume as they are all summed together. This is where a small subtle amount of compression really brings the piece out and allows more quiet details to be heard but without noticeably damaging the dynamics.
best thing to do is load up a compression vst, have a play around with the settings at extreme values to see what overcompression sounds like and get a feel for it, then just tweak the settings until the compression is not noticeable but the improvement in volume and general sound is.
For my music, which is fairly poppy, i tend to use a bit of compression on the drums (drums can take a lot without sounding bad!) and then a bit of compression over the whole track. it really depends on the tune and the type of material but there is almost always a reason to use it, just be careful not to go overboard! i like using reverb before compression so the compression brings up the reverb and makes everything sound 'bigger', some people use compression almost like an effect with huge compression on individual drum hits for that big squashy sound ( chris clark/clark on warp records has lots of tracks with such sounds).

i've gone on long enough! - have a fiddle, if it sounds good, use it! is pretty useful (and it goes into gain and ratio which i actually should have mentioned but it puts it more succintly than i would have!)
also check out this: which shows use of compression on records over the years and tells you why its so bad to overcompress.


Quote from: "anboi"compression is sort of like an automatic volume control which you can set to make loud bits quieter or quiet bits louder (although i think this is called an expander too, lots of plugs do both).
an expander does the opposite.


2 speed-goddamn-focus
That's right. I still have such equipment for analog tape-recorders.
Recording with compression, playing back with expansion.

One aspect about compression I want to emphasize is the recovery of things.
Zip is a compression method where you expect after unpacking to get the original material.
This is called non-destructive compression. Mp3 or Ogg are formats with destructive compression.
Meaning that you definitely loose data of the original full-byte wav-file, which can't be recovered any more.
This loss is not so much in the sound-color area of 'quality', but more in the dynamics area, as mentioned.
An alternative for this is the Flac-format for audio, non-destructive, so you can reproduce the original wav.


2 Sam Zen
There's compression and there's compression. Since KrazyKatz is asking about a "compression VST" perhaps we shouldn't start talking about file-compression. That's an entirely different thing from a dynamics compressor.


Good point. Although there still are connections in the 'dynamics' aspect, besides the filesize aspect.


Quote from: "Sam_Zen"Good point. Although there still are connections in the 'dynamics' aspect, besides the filesize aspect.
You still have the the same dynamic range, it would be wrong of a file compressor to change that. We're probably talking different "dynamics"


Probably. But while conversion to mp3 is primary used for file compression, at the same time this has severe consequences for the dynamics of the song. Load a heavy-compressed file in a wav-editor and you will see a big rectangle on your screen. No subtle dynamics whatsoever left.