1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

What does a compressor and limiter really do?

Discussion in 'Recording' started by phulden, Aug 23, 2004.

  1. phulden

    phulden Guest

    I thought I understood the concept of a compressor, but the more I think of it the less I understand.

    A compressor looks at the music to see whether it passes a certain limit in amplitude ("volume"). The limit is called the threshold.

    If the music passes the threshold the compressor pushes the amplitude down by an amount as given by the ratio. Compression can start directly of after a while by increasing the attack time. Compression can remain even though the music has gone below the threshold by incerasing the release time.

    (Then there's the concept with hard/soft knees and sidechaing, make-up gain and enhancement, but I think I understand all that...)

    1. Does the whole amplitude get compressed at the peak or just the part above the threshold (the latter should end up as distortion shouldn't it?)
    2. What is the difference between peak compression and RMS compression

    I've been told that a limiter is essentially a compressor with a ratio of 1:infinity. But in my mind that would produce a clipped peak at the threshold level and it wouldn't sound good at all.

    3. How does a limiter really work?
    4. When (and why) should I use a limiter? During tracking, mixing, mastering or all of the above?
  2. jonyoung

    jonyoung Well-Known Member

    Compressed or limited signals beyond the threshold aren't necessarily distorted. The idea is to set the threshold and ratio in such a way as to prevent distortion from happening in the first place. If a signal has a wider dynamic range than your system is capable of producing, you have to determine the right combination of settings to keep the signal from clipping and not hear any pumping or breathing from the comp (unless you want that as an effect).
  3. Screws

    Screws Active Member

    The entire signal gets compressed beyond the threshold. Remember that all compression is doing is turning down the volume. If you ever rode the faders during a take or a mix, you were the compressor.

    The attack is how quickly you pull it down, while the release is how soon you bring the fader back to its original place.

    The ratio is basically how radically you react in pulling down the fader as the signal exceeds the threshold. In other words, how far you pull it down. A ratio of 2:1 means for every 2db of gain increase the compressor only allows 1 db of increase. This follows for 3:1, 4:1 and so on.

    A limiter means your ratio is so large (generally 10:1 or more) that you are almost completely limiting the top end of how loud the signal will get. 10:1 means that the original signal has to go up 10db before the compressor allows a 1db increase. So with a 10:1 ratio, you would need a 50db increase above the threshold to end up with an actual 5db increase.

    Infinity:1 means the gain will not go any further than the limit you have set, period. even if you are close miking a 747 on takeoff the gain will not increase more than 1db.

    Hope this helps, and encourages you to ride the gain sometimes, since a good engineer with musical sense and good ears is always a much better and more intelligent compressor than anything you can put in a rack. After all, an electronic device during mixdown doesn't know that the singer stepped back from the mic a few inches too much on the chorus.

    Hope this helps.
  4. phulden

    phulden Guest

    Thanks Screws, good explanation. I get it now.

    So I got that sligthly wrong then. I thought the attack was the time the compressor waits until it compresses and the release how long it still compresses. But what you're saying is that the compression is gradually increasing during the attack time and vice versa during the release time?

    Well, so when do I need the limiter, it seems to me that limiting is kind of a drastic thing to do to a signal (well I know it's used a lot during mastering - but during tracking or mixing it seems quite radical and un-natural sounding to use that much compression)? And still - what is RMS compression and when should I use it (I know what RMS is being an engineer...)?
  5. Screws

    Screws Active Member

    No, you were right. I wrote it wrong. The attack is how soon after the signal exceeds the threshold the comp kicks in, and vice versa for release.

    I was trying to give a correlation to lowering volume by hand.

Share This Page