The compression question thread

okay, maybe this is totally related to the fact that i learnt how to compress things properly using an 1176 - which has the most useless numbering schemel not to mention attack and release knobs which go the 'wrong' way ; but i totally pay not attention whatsoever to numbers on compressors, and turn knobs til it sounds like its doing what i want; be that smashing, taming, smacking , popping or glueing
 
okay, maybe this is totally related to the fact that i learnt how to compress things properly using an 1176 - which has the most useless numbering schemel not to mention attack and release knobs which go the 'wrong' way ; but i totally pay not attention whatsoever to numbers on compressors, and turn knobs til it sounds like its doing what i want; be that smashing, taming, smacking , popping or glueing

i kind'a apply that after getting the numbers to work... it's like first i do what i should, then play a TINY bit till it sounds ok
 
i kind'a apply that after getting the numbers to work... it's like first i do what i should, then play a TINY bit till it sounds ok

Yeah.. I couldn't stand a plugin that had no numbers, just using my ears. The numbers get me in the ballpark because I know what gives me what I want, and then I can tweak with my ears from there.
 
okay, maybe this is totally related to the fact that i learnt how to compress things properly using an 1176 - which has the most useless numbering schemel not to mention attack and release knobs which go the 'wrong' way ; but i totally pay not attention whatsoever to numbers on compressors, and turn knobs til it sounds like its doing what i want; be that smashing, taming, smacking , popping or glueing

Yeah, the attack/release settings on the 1176 are really counter intuitive. Not to mention the "hidden" all buttons ration setting (which I guess isn't really hidden on some versions of the 1176 and some plugs).

Another thing a lot of people don't realize about the 1176 is that when you switch the ratio to 12:1 or 20:1 the internal threshold is raised, which makes the unit behave more like a limiter.
 
One thing I´m still wondering about is whether to set the threshold lower to achieve more gain reduction or to use a higher ratio in order to do that.
As I saw the CLA videos posted somewhere here I was surprised how much dbs he took off. I tried it on my last two mixes and I really liked it with certain compressors.
Before that I was quite careful when setting the threshold too low but if you have a good sounding compressor it can turn out quite well.
 
One thing I´m still wondering about is whether to set the threshold lower to achieve more gain reduction or to use a higher ratio in order to do that.
As I saw the CLA videos posted somewhere here I was surprised how much dbs he took off. I tried it on my last two mixes and I really liked it with certain compressors.
Before that I was quite careful when setting the threshold too low but if you have a good sounding compressor it can turn out quite well.

depends what you want.
sometimes a higher ratio with a lower threshold can sound better than a lower ratio with a higher threshold. Ya dig?
 
One thing I´m still wondering about is whether to set the threshold lower to achieve more gain reduction or to use a higher ratio in order to do that.
As I saw the CLA videos posted somewhere here I was surprised how much dbs he took off. I tried it on my last two mixes and I really liked it with certain compressors.
Before that I was quite careful when setting the threshold too low but if you have a good sounding compressor it can turn out quite well.

If you're looking for more gain reduction, it's easier to achieve by lowering the threshold (as long as you have at least a moderate ratio setting). The ratio only affects the signal when it surpasses the threshold in the first place, and a higher ratio will act more severely upon the audio when it surpasses the threshold. But if you lower the threshold, more of the signal will surpass the threshold in the first place, giving you more opportunity for gain reduction and a more compressed sound. Does that make sense?
 
The problematic situation around understanding compression I think is found in the issue of being able to understand what you hear, and being able to translate it to graphs and other visual contexts and vice versa.

You need to be able to with your hands feel the way you are sculpturing the sound, not just knowing the facts behind it and fumble in deafness.

It's very easy to screw up the mix with compression, if used wrong, but with the right mind behind it it could clean up all the trash and bring forth the real goodies that makes you smile like
jim-carrey_1.jpg


Rule no.1 is to NEVER use compression "just becouse everyone else does", you have to be as humble to the compressor as to your firstborn child mate! otherwise you will turn them both(the child and the compressor) into pitbulls that might one day crush(squash?) your house down to the ground cuz of too high ratio and too low treshold with fucked up attack and release times.

The way sound is treated often nowdays could sound ridiculous on the paper, you compress the drums for smack and attitude on individual/drum buss and then squash them back in line on the master with limiters, but if this is done right, it still retains the smack/attitude but with very good volume without hitting red.

There are a ton of reasons it either sounds right or wrong, don't loose your patience, this needs a LOT of experimenting and experience to get right but it is definently something you need to know how to handle if you wish to mix with confidence. Fact: if you think compression sucks, it's you that sucks, not the compression by nature.

I'll line some insights i've made during my path (trying) to master engineering, which is not even close to reach accomplishment btw.

·Several stages of compression is often needed for desireable results, sometimes my tracks is going throught 8-9 compressors lined up after each other, which brings forth the real magic and attitude of the tracks.

·When compressing individual instruments, don't freak out if it sounds fucked up in individual, when i mix death/black metal vocals it often sounds as the singer was on crack just run the marathon 10 times before hitting tape. But when fighting with the whole mix throught a(or several) mixbus comp(s) and limiter it sits there like fucking on LSD.

·Be sure that it sounds friggin amazing(EQ wise) BEFORE hitting compression, as the compressor would react strange if for an example hit with a very bass heavy track, EQ'ed POST compression. Much ppl apply the EQ-comp-EQ rule, making the sound neutral before comp, and freaking out with attitude POST-comp.

·If you don't hear/understand what you do, STOP, rewind, re-learn, redo, do right.

Ohh god I wet my pants every time I hear good compression...

Cheers:Shedevil:
 
P.s Waves CLA=win

Agreed. In fact, these are some of the easiest units to get compressing with. Most early/plain digital comps have an array of features all that take their tone from being 'shit' to 'abhorrent'. The LA3 and LA2s with their reduction and output knobs are a god-send in many applications. Likewise the 1176 with its 'slam it fucking hard' to 'bring rapture upon thy audio' range is easy in itself. Attack ranges from a few microseconds to 1ms, so you're still compressing fast no matter what's going on. You get to know their character and you're sorted. It's such a musical approach to mixing, and one that's much needed in today's very sterile, digital environment.

As the man behind Kush audio said... compression is one of the hardest things to get right, and one of the easiest to screw up. You can know all the concepts, but still not know how to incorporate them into your process when you mix.

To think of the compressors musically, imagine it this way:

The Attack controls the thickness of your transient. It controls how strongly your drums punch, how much breath pop comes through on a singer, and how up-front the bass guitar will sound. Reduce the attack to send tracks 'into the back' or keep them 'in the pocket'. Slower attack is looser, but punchier.

The Release controls how quickly the compressor 'recovers'. Quick releases can be transparent, but too quick under certain settings will cause your audio to go hopping around like a 'roided pogo stick champ. Good starting point is to let the release complete it's cycle just before the next 'hit' or phrase of audio comes through. Keep it musical, see if you can get the pumping to be in time with the music.

The ratio basically controls how brutal the compressor is with your audio. If you want to take off a lot of dBs, then be gentle and keep this thing between 2:1 and 4:1. Higher ratios are for limiter-like compression where you just want to shave the peaks.

The threshold is fairly self explanatory. Most comps I use these days have this tied into the input gain, so it can even be overlooked as a control.


Now, there are other controls which can complicate things on some compressors. Controls like the 'knee', which alter how the compressor responds as the audio nears the threshold. Then there are the program-dependent compressors which will adjust their attacks, ratios and releases based on what you're feeding them. Designed well, these can be some of the most musical out there. The LA2 and LA3 are great examples of this. Their ratio is in effect variable, and their release depends on how strong the transient feeding them is. A fast, sudden transient will cause a fast, sudden release. A 2-stage release, at that. It's said that the first 50% of the release happens after 60ms on an LA-2, and the rest happens over a few seconds, depending on the GR. Sweet.

Anyway, the practical effect of all this only becomes evident when you dabble with these kinds of compressors.

I'm all about workflow and simplicity, so the LA2 and 3 models work great for me. I want to pull up my 'tone', being the box itself, and just choose 'how much'. The in-line SSL Channel comps are great too. Program dependent slow attack, a 1ms fast attack, with threshold/input gain tied together, auto-adjustment of output volume, and a ratio control. About as flexible as I need it to be. I don't want to be messing about with the side-chain filter, the knee, the release characteristic, 'thrust' and whatever else. It is the designer's job to create a tool that will sound great on a variety of sources, and it's my job to decide how much of that characteristic is needed on my tracks. Those who provide an array of options - many of them sounding purely bad - are simply just copping out.
 
I have 3 questions.
1.In the master bus why you would apply compression? If so, what kind of compression you usually use? is a fast attack comp to smash the peaks? or a slow attack to make the transient more noticeable? (i know it depends on what you try to achieve, but maybe you can give some examples to clear the point)
2.The fast attack comp is used to flat a little the mix and give a little more volume. But in that case, are the snare and kick transients affected? or the volume applied counter acts the smashed transients?.
3. To have a more percussive guitar tone, is it valid to compress it with slow attack and fast release? so the first transients of the notes are louder than the tail?.
I know maybe are some basic questions but it's like to get the knowledge given on this thread tight.
Thanks
 
Thanks for bringing this thread up mephisto, I was searching this. :D need to do some reading...


I understand compression in principle, but what I can't understand about compression is that when I want to even out the dynamics of a snare, for example, so it doesn't disappear under the mix whenever there's changes in the playing dynamics, WHY can't I hear even a subtle difference in the sound no matter what parameters I tweak? To my understanding, I've tried setting it to ultimate-squash-your-fucking-mother but NO! There is no change. The snare just keeps coming up and down in the mix. What the fuck?

It's pretty frustrating when you just don't get it. I'm hoping reading this thread will help me some more.
 
Ok here are some questions about parallell compression
A) what is its purpose?
B) what is it most suited to
C) when you're slamming the copies for parallell compression do you creat a high frequency copy and a low frequency copy?
D) how can you tell if it sounds better? it usually increases the voulme so to me tricks me into thinking its better

and something slightly important THIS IS NOT SIDECHAINING!
I saw it referred to as that in soundonsound magazine, just dont want people to be misinformed