There are two main ways of using a compressor - there's "problem solving", in which you want to achieve a particular solution, and "creative" use, which involves radically changing a sound, and generally will utilise some fairly extreme settings. We'll come to creative use later - but first, let's stick with those extreme settings for a minute.
Most people know the basics of what the various things on a compressor do, so we won't dwell; attack is how long it takes to kick in, release is how long it takes to reset back to zero, threshold tells it when to start working, and ratio by how much. But it's not always easy to figure what this actually means in the real world, so here's a useful trick when putting a compressor on your channel. Set it up initially with a high ratio (let's say 15:1 or more) and then bring down the threshold so that you're taking off a large amount of volume - maybe 20dB.
This will make your channel sound like a horrible mush. So don't leave it like that. However, it will also dramatically emphasise the effect of the attack and release. Change these settings and you'll see that a long release (hundreds of ms) will track slower volume changes - like the difference between your singer's loud chorus and quieter verse รข€“ while a short release will reset rapidly between notes, evening out the level of individual hits, but leaving the larger scale stuff untouched.
At the other end, a short attack (i.e. 10ms) will cut things off sharply like a limiter, making your drums (for example) able to achieve a higher overall level, albeit at the expense of a lively, dynamic sound. A longer attack (i.e. 40 - 50 ms) will allow a transient through but may squash the body of the sound; in this way you can control the overall level while retaining a certain natural feel. Alternatively, you can use this setting to emphasise transients.
Listening to the channel like this should give you a clear view on what the compressor is really doing. So once you've tried various attack and release settings, and they're now doing exactly what you want, back off the threshold to a more moderate level. "Moderate level" is, of course, relative and depends on what you're processing, but you should alter it in conjunction with the ratio.For something that needs to feel kind of natural, like a vocal, acoustic guitar or even an old soul sample, then it's important that your compressor isn't suddenly yanking down the volume. So a low ratio (like 1.5:1, or even less) coupled with a low threshold (such that the compressor is removing around 3dB) will mean the compressor is triggered often, and consistently nudges the volume down. This will feel more natural than a higher threshold with higher ratio; the overall effect may still be that the volume ducks by 3dB, but if it happens abruptly then the effect may not be very musical, and unsuited to a more delicate sound. It might sound great on a bass or drum track though; something with a bit more energy.
People often wonder where in the signal chain the compressor should go. As always, it depends, but as a general rule of thumb any 'fixing' should go before the compressor. For instance if you've got some hiss on your channel, then a compressor can make this sound louder - so try to remove it with EQ, gate or noise reduction before it gets to the compressor. On the other hand, if you want to boost part of your sound with an EQ, then you might find that the part you've boosted is now triggering the compressor, and being turned straight back down again. So in this case the EQ should go later.
It may sound advanced, but don't be afraid of having two compressors one after the other. If you feel you want to to tame some peaks, but also control the overall level of a part, it's no crime to have a compressor with a short attack and release followed by one with longer settings. But be careful - the more compression you use, the easier it is to accidentally squash the life out of a sound!Creative use is a completely different kettle of fish, and you can pretty much discard the rule book. For a good example of the "extreme" settings mentioned earlier, take a listen to someone like Flying Lotus or Rustie. Check out the way pads, FX and hiss drop suddenly out of the mix and then surge back - this is created by having low thresholds, lopping off 15 or 20 dB with a heavy ratio and a long release of 200 - 400ms. For a soft sound like a pad, this can also be triggered with a sidechain input, for instance the kick or snare. Or, you may even set up a channel that has no sound output, and its only purpose is to trigger a sidechain.
Compressors driven hard like this can change the sound dramatically - try pushing a sub-bass heavily through a compressor or two, and you'll hear it create higher frequency harmonics that weren't there before, as it tries to keep up. You can also try some sound design effects - layer up some delays and reverbs on a sound, then squash them hard with heavy compression. Do this a couple of times and you'll end up with a shifting pad or drone sound full of atmosphere.
As we mentioned at the start, compressors can be confusing, and difficult to use correctly. But if you follow some of the tips above, and make sure you're definitely improving the situation with some careful A/Bing, you should find the fog starts to clear a little. So fire up that plugin and get listening!
