Compression

[lem]

Leading on from the 'warm bass' thread, I started thinking about compression....

When I began production I was reading about compression and had this naive idea that it was responsible for making music sound a certain way. It always seemed a bit cryptic, and in the early years I was using it in completely the wrong way. Now, after years of playing around with it, I am able to use it to achieve the results I want...But I still find, unless I am trying to level out a live performance, the signal sounds a lot worse with a compressor on it.

It has been mentioned many times about compressing a kickdrum to 'fatten' it up. Maybe I have the wrong idea of 'fat' but to me it just sounds loads worse...
I do use it on busses and auxs, but never to add a sonic character...

My question to you all;
Does anyone really use compression? If so, how do you use it?

[AK]

If I use it I tend to use small amounts but might use a fair few compressors.

I find that if I'm going to use them, the effect I get with just a single compressor over say, a drum buss, can be too destroying of the velocity levels and other aspects of the sounds, meaning that either the compressor is simply just touching the highest peaks in that buss and isn't really doing anything worthy, OR, if I use it a bit harder to reach levels below that, the peaking bits ( like kicks/snare etc ) lose a lot of impact, depth and are otherwise altered somewhat from their original sound.

With small amounts, I can use a setting with say a long attack inserted at say a kick and say a snare too. Depending on the sound, I might set that inserted compressor to say a really long attack, like 100ms, and along relase with a really low threshold like -36db, then a low ratio, say 2:1. Then with the make up gain, I will bring the kick back up to its initial level. I tend to use this type of setting when inserting a compressor on a kick or a snare when I want to increase the snappiness of the attack transient without reducing it to a click and squeezing the life out of it or making the compressor work too hard. The ratio setting can be increased or decreased while leaving the other settings untouched and has an effect of bringing out the attack and reducing the decay the higher the ratio. But leaving the ratio low will cause the effect to be subtle just pushing the transient through the mix a bit more.

That's at the source sounds - which would obviously vary somewhat. Over the buss, I'll use another compressor but this time with different settings, maybe something like a 4:1 ratio, a 10-13ms attack time and say a 50-100ms release with a threshold set to somewhere around -10db. This time, I'd be looking not to shape individual sounds but just to keep the peaks in check from all the drum sounds and to form a sort of glue in the consistency of the dynamic activity - which kinda helps all the drum sounds breathe as a single entity. Here, I would usually just experiment with the threshold setting then, leaving the other settings untouched, I'd take the threshold from -10db or wherever and drop it to zero so having no effect. Then gradually pulling the threshold back and looking at the gain reduction meter and using your ears, you can start to see/hear the effect it's having over the buss.

That said, I'm not really big on the use of compression, I used to use it a lot before but I wasn't that much into bothering about velocity settings and envelope shaping - which I find gives me workable results. If I'm going to use it though, like I say, it's usually subtle, I don't like hearing the effects of compression but find it helpfull sometimes when it comes to transients/punchiness etc but for general purpose stuff, I usually use one 'open airy' setting and another compressor which catches peaks and squeezes a little.

tried sidechaining, NY/Parallel compression and lots of things in the past - as you do. But I don't like or need sidechaining for what I tend to do with bass and I find NY compression messy and not worth it for dance type stuff where the dynamic range isn't that big. I read it was a thing used on classical type music where they wanted to preserve the huge dynamic range classical and orchestral music has, yet bring about a lot more power to certain attack portions of sound. You can understand how this would be useful in that context. A compressor on a send or with a dry/wet knob could be given extreme settings which would just sound awful by itself but adding a small amount of that wet signal into the dry mix would give a greater sense of depth and presence.

Some of the software I have used acts differently to the next. It's not like you can replicate a setting on one compressor to another as you often get totally different results. I think if you're gonna use it just experiment with as much as you can but I always find the results better when using a few compressors doing small amounts rather than a single compressor trying to work too hard.

[::BLM::]

I use tiny bits of compression on my kicks for example to tighten them up.

other then that im not really using it.

[steevio]

i dont use it at all now.
but like i said in the other thread i would use a top of the range analogue compressor if i could afford one for the odd job.

i just make sure my levels and envelope shapes are perfect for the track to groove right, and if i need a bit of saturation i do it by slightly over-driving the channel on my desk.

thats all the jobs of compression done right there in much more controllled and less crude way.

for mastering its different, very slight compression can tighten up a track, but i am much more likely to use slight limiting to pull back wayward peaks, so my main grooves arent affected too much

my motto is get it right from the start - with everything

[deccard]

to hear what a comp does its best to adjust the output level to the original signal level. so when you bypass it you hear exactly what it does.
i use compression always while tracking to tame the peaks. i have a good hardware comp which is easy to use (jdk r22 has auto attack/release) for that purpose.
ITB i mostly use compression for envelope shaping or on the masterbus. for shaping i have my plugs i tried and work best for me and for the 2 bus i use different ones.
i really like comps with wet/dry knobs so i can use parallel compression without using another bus. very handy option. especially on the 2bus.
the last year i liked to mix into a comp on the 2 bus. didnt do that for some years. my go to plugs are api2500 and the glue.
the api has a specific sound that only fits 30% but when it fits its awesome. so more often i use the glue. i can mix in the dry signal on the glue and can get a very good sounding tight mixdown if i want to. when i mix into the comp i put it on when i already got some basic track idea. over time i adjust the settings or you can easily end up mixing against the comp instead that it helps you.
finally i run my stuff through my outboard comp and my culture vulture. i test how my mix sounds without the 2bus plugin and using the hardware comp instead. actually in 50% of the mixes i turn the plug in off. which is an interesting as the jdk only has attack release automation. maybe i need to adjust the mix a bit but still get a better more open result with the same punch. maybe i just like the jdk sound (jdk is owned by api). also i dont use high ratios on the 2bus. like 1:4 would be an exception.
so the mastering dudes just have to find resonance frequencys and get that thing loud as the other stuff without destroying my mix. (not so easy to find good mastering guys as nobody wants to pay for quality.)

btw i never compress kicks directly. eq yes and the kick gets some compression when i use a comp on the 2bus.

[simonb]

to hear what a comp does its best to adjust the output level to the original signal level. so when you bypass it you hear exactly what it does.

This is great advice, particularly for compression but for any type of audio effect. The volume can have a big influence over whether you perceive something as sounding better or not so if you compensate for this it eliminates that variable and actually lets you hear the difference. I remember reading through a massive thread on the Reaper forums with some guy spilling all his mixing knowledge (maybe one or two of you have read it...) and that was one of the main points he kept drumming home - listen at a sensible level and match effected/uneffected levels for comparing.

[NoAffiliation]

if the compression sounds bad it's probably overdone.

thin sound = bad relationship to peaks and average. speakers are powered by amplifiers which only have so much energy. 6db is roughly = to a 200% increase in perceived volume. now think about if your kick drum had a peak that was 6db louder than the actual body of the sound. keep in mind peaks are extremely short (thin), nonetheless the amplifier will use all of it's energy to reproduce this peak rather than the body of the sound, resulting in the perception of a thin flat sound. it's just energy, once the amp has worked this hard to reproduce this it can't recover in time with enough energy to give the meat of the kick what it deserves.

what you want to do is get rid of the peak without fucking up the RMS. just watch the meters and compress until you start to see the average level drop, if you are really careful the compression will be transparent and you won't notice it switching the compressor on or off.

most likely you could compress every channel in your project by 3db without noticing anything.

so 6db is a 200% difference in actual loudness. if you transparently chopped 2-3db off every channel in your project think how much louder your track is going to be. welcome to loudness wars

i prefer using the free smartelectronix osc scope when compressing, alot faster than watching meters and you can see what the compressor is doing to the envelope

[lem]

I'm pretty sure I get how to use compression. Its the idea bringing out the transient of a kick with a slow attack etc...
There are other means to do so which, I think, are a lot more musical...