This comes from the book 'Mixing with your mind' from Mike Stavrou. Is the best text I have ever read on compression:
This chapter about compressors of the book "Mixing with your mind" by Mike Starvou is highly regarded and praised to learn how to use a comp. Here it is:
It's Like Cracking a Safe
Compressors have four basic knobs (parameters) and the key to classy compression is as simple as the order in which you reach out and focus on adjusting those knobs. When you get the sequence right, you'll hear more clearly the effect of each parameter - thereby arriving at a truer and more musical setting.
The compressor's combination lock has four tumblers. Adjusting them in a special order also prevents you from returning to a previously adjusted control. Don't you hate it when you are happy with the Release time until you fiddle with the Attack? They affect each other when adjusted randomly or out-of-sequence. Chasing your tail is about to become a thing of the past. Like cracking any combination lock, once a tumbler falls into place, you need not return to it. Each step represents decisive progress.
Getting started (temporary settings)
To crack this combination, you will need to set all the controls to a temporary setting while you focus on one parameter at a time. Once the first one is set, that tumbler falls in place, leaving three more to crack. Focus on the next one - listen - adjust - and tumbler number two falls into place and so forth. Approach this safe-cracking exercise in a different order and you will arrive at a different result.
* Attack to anywhere
* Release to minimum
* Ratio to maximum
* Threshold to sensitive
1. Attack
The first thing you do is set the ratio to as high as it'll go - 20:1, infinity... the highest you've got. Next set the release time to as fast as it'll go - which, admittedly, is faster than you'd ever want it. Then, drive the audio into the unit, either by lowering the Threshold or increasing the input (depends on the unit), and listen while you adjust the only the Attack time.
Listen to the Attack - the leading edge of the sound - while rolling the Attack knob. Try to ignore the horrible pumping caused by the after effects of the fast Release - just listen to the Attack. (The ultra-fast Release lets you hear far more individual attacks than a slow setting.)
Listen to the front edge of the sound. Notice how the Attack knob affects the size of the hit. So, if it's a snare drum that you are compressing, and the Attack is on a fast setting, it's as though the drumstick is really skinny.
Alternatively, if the Attack is on a slow setting, it's as if the stick is much thicker. Likewise, if it's an acoustic guitar and the Attack is on a fast setting, you're just hearing the finger nail come through as it hits the string; while if the Attack is slow, you might get the whole strum through - the entire transient bypasses the compressor. So, forget all the after effects, just listen to the thickness of the Attack until it's "tasty" - you might want it thin, you might want it thick, just think aesthetics. And then, because the ratio is so high and the release is so fast, you'll be able to hear the affect of the Attack time much clearer than if they were on any other setting. This technique effectively "turns your ears up" to heighten your perception of the Attack time control.
2. Release
The second step is to play with the Release time. "Release" controls the speed at which the sound glides back at you after being punched away. The trick is to get that speed to become a musical component of the sound. You might ask, "Do you mean in time with the music?" or "With fast music do I set faster than I would for a slow ballad?" Perhaps, but certainly don't think, "I want it fast because I want to compress the crap out of this" - don't do that. In fact, make it as slow as you can, so the compression envelope bounces back to reinforce or establish the groove of the music. Remember, any dynamic movement in a song affects the groove, and compressor/limiters are no exception. (Whether the Singer is moving back and forth from their mic, or you're madly wiggling a fader, or a compressor is pushing and pulling on a sound, the groove is at risk of being enhanced or destroyed by dynamic movement.) So, don't set your Release to a fast setting just because you want to hear something buried behind the sound. Forget that. There are bigger fish to fry. You're already compressing a little bit, so the background sounds will come forward anyway. Instead, you want to think, "How slow can I get it while maintaining some control?", because the power in the groove is really a slower-moving, subliminal yet powerful wave - it's not an ultra-fast thing that's there to crunch your sound. Even in a frantically fast-paced tune, a slower, subliminal undercurrent carries most of the power. For example, you might have it so slow by the time the next hit comes along it's not quite fully released. But that's okay. A formulaic approach might intellectually tell you that it has to be fully released before the next hit, but that's more math and less groove.
Listen to the Release. Feel the way it glides or bounces back at you and there will be a point where you sense this bounce-back is kind of like a swing -almost like someone is swinging from a rope in a tyre in groove with the tune. It doesn't have to be perfectly in time, because a groove - as anyone who teaches music will tell you - should keep time, but not necessarily play the time.
Never play the metronome. Never play the conductor's baton. So, don't just make it a quarter of a beat or whatever, just look for that groove, and that's your release time. Make the rush of the Release a musical component that pushes you into the next beat without pre-empting the beat. Let the musician hit you while the pressure is still rising instead of letting the compressor finish its swing - dead air - lifeless moment... no good, Allow the compressor to push the sound towards you until the music makes it's next statement.
If, however, all you care about is maximum volume (no matter how detrimental to the groove that might be), then ignore this last paragraph and set the Release to "maximum irritation"! But I must add that if you aim to make the product likeable (extremely groovy, for example), the wrist of the listener will always turn up the volume for you more effectively than any brick wall compression ever could.
3. Ratio
At this point, the Ratio is set to maximum, so it's going to sound over compressed. So the next job is to take the Ratio and lower it as much as you can without losing the effects you created with your Attack and Release settings.
Think of the Ratio control a bit like a telephoto lens - the higher the Ratio, the smaller the sound is - although it will be more controlled. The lower the Ratio - as in 2:1(given the same output voltage), aesthetically feels like a larger image. So, the lower the Ratio the bigger it is - but at the risk of getting out of control. Meanwhile, the higher the Ratio, the smaller it is - although more contained. The idea is usually to try and make it sound big, but in control. So, bring down the Ratio, then when you don't hear the effects that you like - the thickness of the stick, the groove you created with the Release time - you can raise the Ratio a little, all the time focussing on size. At this stage, don't think about Ratio in terms of numbers - just about size and firmness of the sound. You know how I often talk about "firmness' and "Hardness Factors"? Well, as you raise the Ratio, the sound will become firmer (and smaller) as as you lower the Ratio it becomes softer(but bigger). So you might want to think along the lines of: "How firm do I want this?"
4. Threshold
The last thing you adjust is the Threshold. It's important to turn the Threshold knob so that it's not compressing all the time. The right setting will see dynamic movement coming to rest at special moments - otherwise you get a flatter, more lifeless sound.
Having uncompressed sound emerging from the processor at appropriate musical moments adds colour and contrast to the sound. For example, permitting the dynamic movement to come to rest in some quieter moments allows that moment to attain a momentary, bigger, 1:1 presence, and prevents it from rushing towards the listener with unwanted noise. It's sad enough that the little quiet moments are small without being squashed smaller still due to high compression ratios. Each time the sound comes up for air, so to speak it attains a sense of reality - a 1:1 ratio.
WARNING!
Most engineers do not realise that Ratios are multiplicative, not additive. If you compress your mix 10:1 and then the mastering engineer compresses it at 10:1 you effectively achieve, not a 20:1 but a 100:1 texture. Ouch! Consider yourself warned. This applies to all compression. If you compress a voice during recording at 10:1 and then in the mix again at 4:1 you don't get 14:1 but 40:1. Next time you mix consider the ratios likely to be used at the radio stations that provide the finishing touch. Ask yourself, "How small a sound can I bear to hear On the Air?"
That Very Expensive Sound
If you follow these steps, set your compressor to the settings in the illustrations, and follow the path of the Yellow Knob Road, then by the time you get to this point in the article you'll have a big and bouncy, firm but flexible, juicy and slippery groovy sound. Or as some would say, "a more expensive sound".
Mike Stavrou
