Compression question

[brevity]

I've been recording for several years, but never really had much of an opportunity to buy nice gear. Now I'm looking into getting into more of a "nice" setup, and want to eventuall get some compression going... now on to the question:

Does the compression go in the chain INTO the recording, as in on each track, or is it all in post like through a send and return? Also, a compressor acts as a limiter as well, right? That is, if I play very loudly into a mic or scream with the gain turned up, it will prevent clipping (assuming I have my limit levels set correctly) of the waveform, yes?

Now, also... if there is clipping allowed, is it symmetrical or asymmetrical clipping?

 

[peopleperson]

Use a compressor when you feel like you need one. You know, on stuff with a lot of dynamic range. As usual, there's no set rule, but clipping sucks. Go guy an RNC. It's the best you'll get for 200 buckers.

 

[xstatic]

I never track with a compressor for dynamic control (clipping, varied levels etc...) Thats what mixdown is for and proper level setting. I do however use compressors quite often when tracking for tonal shaping, envelope manipulation etc... If you are clipping your converters, trun down your signal before the converters. Most likely you are pushing your preamps too hard anyhow if clipping is a real issue. The reason I do not care a whole lot for the RNC's that everyone loves is the same as I stated above. The RNC's do well for a little but of clean gain reduction, but the just don't sound good to me when you step on them pretty hard and don't seem to "shape" things well for me. Thats probably why I have a rack of nice vintage compressors that each really impart their own sonic signature. Just my 2 cents

 

[SouthSIDE Glen]

Does the compression go in the chain INTO the recording, as in on each track, or is it all in post like through a send and return?[

It can be used at any point in the signal path. But in general it's used as part of the recording signal chain only when the recording engineer needs it to keep the levels tamed and within a decent recording range; e.g. when a vocalist has no mic technique or when doing a location shoot where the levels vary greatly because of changing environmental conditions.

Otherwise, compression for sonic effect ("getting 'that' sound") or for fitting the mix together is usually reserved for the mixing stage (the first half of "post"). Compression for final polishing of the mixdown is done in ine mastering stage (the second half of "post").

Also, a compressor acts as a limiter as well, right? That is, if I play very loudly into a mic or scream with the gain turned up, it will prevent clipping (assuming I have my limit levels set correctly) of the waveform, yes?

Not necessarily. While a compressor and a limiter are cloesly related cousins, there are important differences. Most compressors are not "brick wall limiters". That is, not only can they not apply as high of a gain reduction ratio as a limiter can, but they often do not have as fast of an attack response time as a hard limiter. In short, most compressors will still let some spikes and transients through and can only compress the rest of the signal by a limited percentage.

If one does not have a true limiter available, a compressor can do the job of a "poor man's limiter" so to speak, in that it will help tame much of the signal, but in the cases where nothing but a hard limiter will do, nothing but a hard limiter will do.

Now, also... if there is clipping allowed, is it symmetrical or asymmetrical clipping?

The clipping is usually symmetrical* unless there is an unwanted DC bias current introduced somewhere along the line that skews the baseline of the waveform off of zero voltage. Such a DC bias will reduce headroom on one side of the AC waveform but not on the other, meaning it's possible to have asymmetrical clipping.

DC bias and asymmetrical clipping caused by DC bias offset are usually not desired. If such a bias is introduced (usually through a poor-quality PC soundcard, if used), it should be filtered or digitally offset before sticking the signal to disc or tape.

*EDIT: Clipping can be asymmetrical on very sharp and short transients where the downstroke of the wave is more rebound than impulse. The actual clip may only happen on one side because the amplitude of the other side falls just a bit short. But this is really still AC symmetry, it's just slight differences in amplitude that determine whether the signal clips or not.

G.

 

[xstatic]

And a compressor will not stop your preamp from clipping even though it may keep your A/D converter from clipping.

 

[brevity]

Thanks guys, I feel like I have a much better understanding of what compressors actually do now... which is exactly what I was looking to gain. Y'all are a big help and I certainly appreciate it. I'd ask what's a good technique to use a compressor, but I'm sure it's just like the rest: you just gotta learn it.

 

[SouthSIDE Glen]

I'd ask what's a good technique to use a compressor, but I'm sure it's just like the rest: you just gotta learn it.

Yeah, unfortunately there's very little "brevity" in the ways of learning compression .

If you want to spend a night or two trying to learn your compressor, I'd suggest starting with one or two individual instrument tracks (maybe a drum, a guitar and a vocal) and them movingto a stereo mix.

Take each track individually and put it through your compressor, starting at a setting of 0dBFS threshold, 0dBFS gain reduction, and attack and release set on either fast or auto.

Move the threshold level down just 2 or 3 of dB, then start turning the gain reduction/ratio and see what you hear. Play with attack and release a bit and see what that does.

Then reset your gain reduction/ratio and your a&r settings, move the threshold down another 2 or 3 dB, and start all over again with the reduction/ratio, listening to changes in sound. Repeat as necessary.

Also, keep your eyes open on these boards for a hyperdoc I'm going to be putting out on compression. The schedule keeps sliding back because of ither obligations, but hopefully in about two weeks or so I'll have something to announce.

G.