Skip to main content

Basic Compression Question from A Newbie

Member for

21 years 3 months
Hi Everyone...

My first post here! Have a very basic question...

I do home recording (for fun) with an Akai DPS16. I have an RNC Compressor and would like to add a touch of compression to the final mix (going to my Philips CD burner).

My music, basically, is in the style of Rubber Soul (I wish!), and I also sometimes record with the bluegrass band I play bass with.

Question...what "Attack" and "Release" settings should I use on the RNC for adding compression to the final mix? Slow/Slow, or Slow/Fast, etc.?

I always try to let my ears tell me. But I'd like to borrow your ears and brains on this.



Member for

20 years 11 months

Stephen Paul Fri, 09/27/2002 - 11:16
Bob... a word or two about compression, especially compressing the whole mix.

Unless you're after an effect, my advice is to stay away from shooting your whole mix through a piece of crap like a compressor.

The fact is that almost without exception, no matter what the claims are, compressors are all much higher in distortion than they are spec'd.

The reason is that when modulation occurs of the entire complex wave's amplitude envelope, the distortion goes shooting out of control.

To quickly give you an idea though, of what these controls do, and the basic difference between compression, say, and limiting, here goes:

Attack time controls the delay between the entrance of the signal into the unit, and the onset of compression. Slower attack times are lower in distortion, and are generally used to increase 'apparent level'. IOW, the overall sound tends to become louder on a subjective basis, because the waveforms are packed together in amplitude, while some transients are getting through.

Generally the faster the attack time, the more transients, which have steeper wave fronts and therefore demand more bandwidth and quicker response to catch, are included in the detector.

As you slow the attack time, the sound will seem to become more 'open' sounding and less choked off, while the apparent level rises and the overall sound thickens.

As you speed up the attack time, less peaks get through and the sound starts to pump more.

Of course this also depends on the release time.

The release time determines the delay for the compressor to return to unity gain after the wave envelope starts to diminish in amplitude.

A 500 millisecond attack time for example, takes a half second to return the voltage controlled amp to the '0' mark.

Just as in the attack situation, the faster the release is set, the more the unit distorts as it tries to follow the hills and valleys of the wave, and the greater the subjective impression of an effect. And the greater the pumping.

Let's talk about ratios for a moment.

A 2:1 compression ration means that if you put a change of two dB into the unit, you get a one dB change out.

If you set a 20:1 ratio, it takes twenty dB of change in, to get a dB of change out.

So the higher the ratio, the more signal is needed for a given change in output levels. And therefore, the more compressed the signal is.

Compressors come in two basic configurations, open loop and closed loop.

Closed loop compressors are by far the most prolific designs.

The difference is that in a closed loop design the output amp level is fed back into the input of the detector, and this is used to control the amount of compression applied to a given signal.

In an open loop system, the detector is on an independent line, and the signal is passed from the input to output, with the detector out of the feedback loop, and controlling the VCA 'externally' so to speak.

The advantage of open loop designs are that you can do things like infinite compression, (infinity to one, meaning that no matter how much input signal you feed in, no additional output comes out of the unit), and the subjective quality of the compressor is a more open sound.

Most closed loop designs cannot do infinite compression, because it would require so much gain to drive the feedback loop, you'd never get there.

One of the reasons I love my LA-2As so much, is that they're open loop designs, plus program dependent release time is part of the basic concept of the design.

Finally, the idea that tube units put out even harmonics and solid state ones put out odd orders, is fallacious.

In fact, even tube designs are predominantly third order harmonics, but the ratio of the harmonic structure is different, and leans toward less third order predominance in the output.

The reason we like second order or in general even order harmonics more, as humans, is because the second order first harmonic is musically related to the fundamental.

IOW, if we feed in 440 Hz, the first harmonic, is 880 Hz, or an octave. Odd order harms are simply not musically related, except at certain intervals.

My advice would be to always also put your compressor in the chain after the EQ. Compressors by their very nature add noise in the release cycle, and putting them after EQ tends to minimize the additional noise, and preserves headroom, since boosting EQ is the same as demanding more output from those amps in the freq. range you're EQ'ing. So the compressor will keep those within limits.

Limiting as opposed to compressing, is really mostly a function of the attack time. Limiters are set up faster to catch peaks, and control overall modulation, compressors are slower and are used to increase apparent loudness and thicken the sound.

There's a whole book I'm sure that could be written about these devices, but as always, use your ears.

If it doesn't sound better, don't use it. If it does, screw the tech data and use it.

I have no idea if any of that helps at all, but maybe it might at least make you a little more inclined to try and get the compression on your tracks themselves right, and preserve the clarity of your mix, unless you're after an effect, in which case, do whatever you want that creates the effect you're after, and to hell with any 'rules'.


Member for

21 years 3 months

archived member Fri, 09/27/2002 - 17:27
Great thread. I have a few questions that follow from your above response Stephen.
1. You advise keeping away from compressing an entire mix. How do you get competitive levels with out compressing (or limiting) the mix? I am not talking about ultra-flattened crappy stuff that is far too commonly heard these days, but more reasonable amounts of compression on the mix. How do you avoid being too quiet (and please don't tell me you leave it for the mastering guy- you must have an opinion on what they are doing to your mix)?
2. I have been using my Valley People 610 to compress final mixes, which I have been preferring toplug-insor the RNC. Should I be using something else?
3. Am I correct in understanding that you compress everything going to tape with the idea that that will be the only compression applied through the entire recording process? Wow.
Thanks for posting such a thought-provoking response. Bob is definitely getting his money's worth on this question. Doc.

Member for

20 years 8 months

RecorderMan Sun, 09/29/2002 - 09:59
first, great response stephan...hang here more often when your schedule allows.

Next, When you print your mixes. Do two versions. Print one with and with-out compression across the bus. This way you'll learn, have your ass covered with an un-processed version, so that a mastering guy can do it better. And if you're compressed version sounds great, but the next day you think you over did it you can have another go a compressing it again.
You're right about needing 2-bus compression to compete. The gods may not approve, but it's a rare mix that doesn't have compression across the 2-bus...rare at leats among the product in stores that is.
You have to learn, so jsut cover your with the compression and print one with out. Slow attacks, fast releses and ratios od 1.5:1 / 2:1 with 1-3db of gain reduction at most will get you in the ball park.