Hey Guys, Recently I've been trying to go from random knob twiddling to more of understanding why knobs do stuff. My first stupid question is:

The Ratio sets the amount of 4:1 means that if an input signal enters compressor and is 4dB above the threshold, it will leave the compressor 1dB above the threshold. So now if a compressor is is getting a signal 8dB (i.e 2x louder then 4dB?), it will leave the compressor at 2 dBs? Is this why a higher ratio makes the sound louder? So if everything was over the threshold the whole signal would squashed by 4 db?

So in order to squash the dynamics of a sound, you should use lower compression ratios???



Thanks, and remember this is stupid question #1... More to come .

I would not put it like "higher ratio makes sound louder".

I would rather say that higher ratio allows you to get a louder sound - depending on other settings - without getting overs(clipping).

The higher ratio the closer you come to what is called limiting. Ratio of 10 and more is considered limiting.

And the higher ratio the greater risk you add unwanted stuff to the music.

Overall a compressor is used for two purposes:

1. To even out irregular strength of signal, say a bass line or vocals.
2. As an effect to color sound

So depending on purpose you select which compressor to use, and what settings to use.

if you set the threshold so low that the signal is always above the threshold, then you wouldn't really do any "squashing". the peaks will still remain the same relative to each other, they'd just all be of lower volume (how much lower depends on the ratio setting).

The threshold is the point at which compression starts, the ratio is how much the signal is reduced by, the attack determines how quickly the compressor starts to work

Yes, so-to-speak. But there's more. THEN you can use the Make-up gain fader on the compressor to bring back up the level to the original (unaffected) signal's level (this is an important thing to understand)(some compressors offer an auto-makeup gain button, but best to get used to adjusting your own levels for desired result). So in effect, you are decreasing the dynamic range between (between means ratio) one part of the compressed signal to the other uncompressed parts of the signal. The uncompressed parts of course being those before the attack you have set and after the release you have set. Remember, attack parameter tells when compression should start and release is how long it takes for signal to reach input level again. And guess what, some of the attack and release characteristics VARY depending on model and design of compressor. Hope it helped a little. As easy a concept as it seems, compression takes a lot of investigating to get one's head around it. It took me some years before I was getting desired results from my own settings but once I did I went crazy with experimentation. I use DMGAudio's Compassion compressor, for instance, and I sometimes get lost in all the nice features, but one can learn MUCH about many types of compression from it. Anyway, all compressors are fun. Good luck!

it a very simple formula/function:

input volume / output volume


Actually, compressors like Maximus give you the ability to freely shape the function freely by drawing a graph. Imo the best approach.

Thanks for the responses guys...

Compressors are actually kind of annoying. I'm still expecting it to behave one way, but when I tweak the knob, it does something different. I'm currently using Ableton Live's compressor, and demo of FF Pro-C. I'm wondering why a synth lines becomes more ballsy, more aggressive it becomes (not distorted but something...)

Thanks for your help so far! Will keep experimenting.

i always use 8:1 cuz 8 is my favorite number

i just find a way to make it work

I think you probably already have a good understanding of what a compressor does. Except for the loudness thing. This is from Wikipedia:



So, basically, the ratio tells the compressor how much to squash the sound above the threshold. It doesn't make the sound louder, it makes the loud parts above the threshold quieter. But, this enables you to turn the sound up louder in order to get a louder average sound.

A comparison between a compressor and limiter is a good way to understand it. Basically, a limiter is a compressor with such a high compression ratio that anything above the threshold is inaudible. You can emulate a limiter by turning the ratio up to its highest possible setting which will usually be infinity.



Just a rule of thumb here...

Aim to not need a compressor. Go back to the question of why you need a compressor in the first place. The reason you apply a compressor to a track is to tame volume swings. If you can tame volume swings at the original source, you are much better off. A very typical example is when they record a melody from a midi keyboard. The midi keyboard will record how hard you hit the keys which will usually result in volume differences between notes. You can usually fix this by manually reducing the velocity differences on the notes. Then, the need for a compressor is much less.

After you've done this, then you can apply subtle compression for a bit of smoothing. But, at the end of the day, if you've already designed your sounds well in your synth of choice, compression is basically pointless and you'll only be changing the dynamic of your synth's sound.

Of course, you might want to squash the mix down a bit at a later stage, but I'd always leave that up to a professional mastering engineer rathing that playing with that yourself.

A compressor is not Exclusively a mix tool It is great fun to employ a comp if one wishes to enhance or dampen a portion within a synth part or beat segments or maybe you want a breathing motion to it. It all really depends a lot on the SOUND you are after. I also use Ableton Live and the native compressor sometimes. Lives comp is just fine for learning compression techniques. IMO you really should avoid putting any money down on various other soft comps before you have a firm grasp of how to use the free one you got with Live. There's been some good instruction here, so speak up if you still don't get it. It's something that has to "click" in the brain from experimenting. Then, creatively speaking, the real fun comes when you learn the various comp models and characters they give to a source. Also perhaps you wish to enhance the tail of a reverb later on. The sky's the limit. Learn Lives comp first however.

PS The well-known SoundOnSound magazine (digital and print) has had great articles on various types and uses for compression through the years. Recommend browsing those articles if you have a subscription. The peopl really know what they are talking about.
Last edited by snigelx on Fri May 04, 2012 3:44 am; edited 1 time in total

Not sure I understand your question here. Try to rephrase it. What are you trying to do?

Nevo said something interesting in a Waves Webinar about compression ratio that I took for myself as well.

He said that a 4:1 ratio is already hard compression and once I took he's advice in usually using between 1.?:1 to 4:1 , my compression is a whole allot better now.

When I want a really hard compression (usually on drums) I'll maybe go for 8 or 10:1 with a hard knee also of course, up for special effects only.

Of course, essentially 10:1 is limiting.

This may seeem a bit pedantic but a more accurate description for the attack is to say that it slows down the rate at which gain reduction is applied rather than how quickly the compressor starts to work as compression starts as soon as the sound crosses the threshold, the attack is used to slow down the rate that the gain reduction occurs. The word quick is better suited to the release function as this determines how quickly the compressor recovers from gain reduction, the release function begins the moment the sound dips in volume not when the sound drops below the threshold as a lot of people seem to believe, the threshold is merely a trigger for the compressor to begin working. Once above the threshold the attack and release are in constant use as the sound rises and dips in volume.