Some Help with compression.

D

d(-_-)b-Phones

New member
I understand that this has prob been asked 1000 times, so here go's 1001.

Compression. Easy word.. Not so easy action for me.

I have been working with compression and EQs lately on everything and I am begining to lose faith. I am using Nuendo with Kjaerhus Gold. It has some set deals in it for kick drum... bass guitar .. vocals etc... But i hate to think that I am genericly following the guidlines set forth by some VST programmer.

I have read countless post / articals on compression. And nothing is matching up. I need a guidline.. I am so tired of the truthful yet cop out answer --"There are no guidlines, No one way." Sure that may be true once you have a good understanding of somthing... but to learn there must be a starting point. I am having a hard time understanding ratios and how they can work for or agants you. I am also having a hardtme with input gain and gain in general. also EQ during compression. all boggles my mind.

Example.

Say I drop the lows of the kick in compression eq to 60hz then gain it up to "sounds decent level" and its dropping -0.2db in compression. at whatever ratio you like ( because i dont understand it ) then when i goto EQ will that same 60hz not effect my EQ like if i have already dropped 60 then i drop another 60 am i dropping 120 does it double? is the eq applied the same?

Please help.

PS -- if your going to say "If it sounds good its good" please just restrain yourself and move along.
 
d(-_-)b-Phones said:
Say I drop the lows of the kick in compression eq to 60hz then gain it up to "sounds decent level" and its dropping -0.2db in compression. at whatever ratio you like ( because i dont understand it ) then when i goto EQ will that same 60hz not effect my EQ like if i have already dropped 60 then i drop another 60 am i dropping 120 does it double? is the eq applied the same?
Click to expand...
I have to be hones, 'Phones, and say that I don't quite understand that question. It appears to relate to how compression and EQ interrelate. It also appears to confuse frequency with decibels.

Let me instead try this, 'Phones. Let me toss our a few basic principles and you can react to what you understand and don't understand about them and maybe we can go from there?

First, regarding the above example: In general (I know you want specifics, but I'm sorry, they just don't exist), you want to EQ before you compress to get rid of unwanted lower-volume stuff that muight otherwise become accentuated by the act of compression. Getting rid of a 60Hz ground hum would be a good example, like you used. Conversely you'd EQ after compression to sweeten the resulting sound to taste. But - again, in general - don't necessarily expect EQ that was used pre-compression to sweeten the sound to sound just as sweet after you compress, because the compression could accentuate frequencies that you didn't EQ.

So, with that in mind, EQ to get rid of "noise", then compress, then EQ to "sweeten". Then, after all the EQing and compression has settled,m you can then set your overall volume levels. to maximum pre-clipping.

As far as compression ratio, those are numbers. How they translate to the real world is not something easy to put a general "if A, then B" rule on. Which is why nobody does that here :). All I can start with is an explanation of what that ratio actually means. If you have a 4:1 compression ratio, it means that for every volume increase of 4dB in the wave form going into the compressor, there will be a volume gain of only 1dB coming out of it. 5:1 means every 5dB gain will come out as only a 1dB gain. and so on. The "threshold" settings just tells the compressor at what volume level to start the compression at (e.g. if your threshold is set at -10dB, that means that the compression will not be applied at any point in the waveform where the volume is less than -10dB.)

Input gain is nothing more than a volume control for the inputs on the compressor. It's kind of like the preamp volume for the compreessor, and has little effect on the actual compression process itself other than ensuring that the volume (voltage) going into the compressor circuitry is at a "comfortable" level for the electronics to work with.

Output gain is just like input gain, except in reverse; it's a volume control for the signal going out the compressor to the next item in the signal chain. This mostly comes in handy when you have fairly moderate-to-heavy compression applied; because the compression lowers the peak volume level and sometimes (but not always) the average volume level of the signal, the Output gain control is often used to boost the overall volume of the signal *after the compressor* to bring it back to a "normal" or desired volume/voltage level.

So, where in all this are hidden the "rules" for how to use a compressor? Nowhere. The "rules" are in the music you're trying to compress. What it takes is an understanding of the relationship between the "flavor" of a given recording and the particular "herbs and spices" that a compressor has domain over to know what to do best. The only way to start that (other than expirimenting) is to take the knowledge of what the "compression ratio" and "threshold" dials actually do (as described above) and apply that knowledge to the sounds and waveforms in front of you. In other words, it takes some thought. It takes, for example, looking at the wave display in the DAW in front of you and seeing just how dynamic the "shape" of the sounds are, looking at the actual levels on the meters and the sides of the wav displays and figuring that if there's one peak at -6dB and another one at -3dB that there's a 3dB difference. Therefore if I set my threshold at -5dB and set the compression to 4:1, the results will be that the first peak will still be around -6dB because it falls under the compressor's current threshold setting (to the more learned readers, we're assumung a hard knee here) and the second will lowered to around -4.25dB.

As you get such a feel for the relationship between what you see on screen and your compressor settings, you'll also be *hearing* how these differences effect the overall sound. And that is what's eventually important. As the Master Men here have shown so well, one should not master with their eyes. However, using your eyes and ears together as part of the learning process for learning just how compression settings relate to what you see and hear will take you a long way to learning how to fly a compressor.

If you can find anything in all the above information that even hints at "rules" or presets or anything like that, I'd like to know what that is. Becaise the BIG variable in thae whole relationship is the waveform; i.e. the nature of the recording your working with. As that changes, the need and application for the compressor changes.

That's not what you want to hear, but that's pretty close to the way it is. Anybody who comes back to you and says "use 2:1 with a threshold of -8 when you are doing this, but 3:1 with a threshold of -10 when you are doing that" is just patronizing you by telling you what you want to hear instead of what you need to know.

Now, given the baseline of information above, what do you need to know?

G.
 
Last edited:
SouthSIDE Glen said:
I have to be hones, 'Phones, and say that I don't quite understand that question. It appears to relate to how compression and EQ interrelate. It also appears to confuse frequency with decibels.
Click to expand...

Thank God it's not just me. I didn't want to say anything because I thought I was even stupider than previously presumed. But I couldn't understand stuff like: "Say I drop the lows of the kick in compression eq to 60hz then gain it up to "sounds decent level" and its dropping -0.2db in compression. at whatever ratio you like ( because i dont understand it ) then when i goto EQ will that same 60hz not effect my EQ like if i have already dropped 60 then i drop another 60 am i dropping 120 does it double? is the eq applied the same?"

I had to lie down after reading that.
 
RAMI said:
Thank God it's not just me...
Click to expand...
Yeah, but I'm giving him the benefit of a doubt and assuming that English is not his first language, in which case I'm not going to hold it against him...I hope that's the case anyway. :)

G.
 
Fair enough...you're a mensch. I wasn't really judging him. It just confused the hell out of me.
 
I was confused too............ but I'm kinda in the same boat as Phones trying to understand why there's not a... "Anybody who comes back to you and says "use 2:1 with a threshold of -8 when you are doing this, but 3:1 with a threshold of -10 when you are doing that" is just patronizing you by telling you what you want to hear instead of what you need to know."...SSG, the whole explanation helped put it in perspective. Thanks.

Thanks for kickin this off, Phones.......... ;)
 
compression is a little confusing to me too. I have'nt used it much but would like to understand it better so that I can use it when I really need to. I have also read 4 or 5 articles about compression and eq.

maybe someone here could help us ignorant folk out? we could post a couple of tracks in need of eq and compression. everyone involved could use the same freely available plugins so that we are all on the same page. the experienced folk could maybe guide us through getting the desired results. does this sound like a good idea?
 
SouthSIDE Glen said:
...EQ to get rid of "noise", then compress, then EQ to "sweeten". Then, after all the EQing and compression has settled,m you can then set your overall volume levels. to maximum pre-clipping.
Click to expand...
This is 3 years worth of advice... :cool:
 
i think that the hard part for me in understanding compression is that I don't know all of the ways in which a compressor can be used. compression seems like a very versatile tool, not just something to squash peeks. can someone give some examples in how compression is used in different situations? i understand what all the knobs do, except for maybe "knee". that's still a little unclear to me.
 
Well, if I have this right, the knee refers to how 'hard' the compression comes in. Say you have it set to come in at -10db at 8:1: without knee set it will suddenly start attenuating the signal at -10db. With knee this is gradual so the compression will start at (say) -15db at a ratio of 2:1 and then increase so that at -8db it will be compressing with the full 8:1 ratio. This should give a much more natural sound and allow more dynamics to come through.

My experience is a bit limited here though, any tips on where knee is useful and shouldn't be used by the more experienced people would be really cool... please.

Disclaimer: Numbers for example only, don't take them as real-life figures.
 
TravisinFlorida said:
maybe someone here could help us ignorant folk out? we could post a couple of tracks in need of eq and compression. everyone involved could use the same freely available plugins so that we are all on the same page. the experienced folk could maybe guide us through getting the desired results. does this sound like a good idea?
Click to expand...

Since there's no one right answer, you'll get as many different results as participants.

I usually try to use the limiter to get rid of the extraneous peaks that don't show up in the waveform on the screen, then apply 2.5:1, -20dB, "auto" for attack and release, normalize before and after to a final mix. For a vocal track, I might crank up the ratio to 3 or 4 to 1.
 
d(-_-)b-Phones said:
I understand that this has prob been asked 1000 times, so here go's 1001.

Compression. Easy word.. Not so easy action for me.

I have been working with compression and EQs lately on everything and I am begining to lose faith. I am using Nuendo with Kjaerhus Gold. It has some set deals in it for kick drum... bass guitar .. vocals etc... But i hate to think that I am genericly following the guidlines set forth by some VST programmer.

I have read countless post / articals on compression. And nothing is matching up. I need a guidline.. I am so tired of the truthful yet cop out answer --"There are no guidlines, No one way." Sure that may be true once you have a good understanding of somthing... but to learn there must be a starting point. I am having a hard time understanding ratios and how they can work for or agants you. I am also having a hardtme with input gain and gain in general. also EQ during compression. all boggles my mind.

Example.

Say I drop the lows of the kick in compression eq to 60hz then gain it up to "sounds decent level" and its dropping -0.2db in compression. at whatever ratio you like ( because i dont understand it ) then when i goto EQ will that same 60hz not effect my EQ like if i have already dropped 60 then i drop another 60 am i dropping 120 does it double? is the eq applied the same?

Please help.

PS -- if your going to say "If it sounds good its good" please just restrain yourself and move along.
Click to expand...


Ratios: 1:1 = for every 1 db of input, there is 1 db of output
2:1 = for every 2 db of input there is 1 db of output.

The compressor turns *down* the volume 1 db for a 2 db input. So, the output is cut in half.

Threshold: = a voltage point that you set to tell the compressor when to start working. A high threshold would be used to "limit" the signal to prevent overloading the input amplifiers of the soundcard (this is called a limiter function). A low threshold is used to compress the signal right from the start (this is used to reduce dynamic range, to make an insrument or vocal "sit" in the mix better by keeping the volume constant all the time).

EQ after compression. The EQ settings affect the compressor and will drive you crazy in a short time if you don't understand what is happening. A compressor "turns down" the volume after a certain threshold value (you set). If you EQ before the compressor, everytime you boost a frequency, you adjust the overall level of the signal. If you are above the compressor threshold, then the compressor turns down the volume again. This is a no-win situation (unless you have a make-up gain stage at the compressor's output). Even then, it is a better solution to set up the EQ after the compressor.


Lastly, VST plug-ins do not act like hardware devices. You do have to experiment more to get a proper sound. What you have to do is read up on what a compressor does and understand why. If you don't know the theory, there is no way to help you because a compressor can do a ton of things and there are no "quick start" settings that will work on everything.
 
Important Correction

To those who have told me they have read and printed out my long-winded post, I finally proof-read it this morning and saw that I had a BLATENT math error in it that would give totally erroneous results. This had to do with my description of applying 4:1 compression to two different volume peaks.

As of about 0930 CDT today (June 24th) I have (hopefully ;) ) corrected the error.

Sorry for the inconvienence, and thanks for all the fish. :p

G.
 
acorec said:
Ratios: 1:1 = for every 1 db of input, there is 1 db of output
2:1 = for every 2 db of input there is 1 db of output.

The compressor turns *down* the volume 1 db for a 2 db input. So, the output is cut in half.
Click to expand...

Actually, the compressor ratio works on linear absolute units and the dB scale is logarithmic based on a ratio to reference value.

The article I linked to has a Part 2 (link at the end of part 1) that shows the transfer functions that might illustrate it.

There are also some links to articles on EQ at the top of the page.
 
acorec said:
EQ after compression. The EQ settings affect the compressor and will drive you crazy in a short time if you don't understand what is happening. A compressor "turns down" the volume after a certain threshold value (you set). If you EQ before the compressor, everytime you boost a frequency, you adjust the overall level of the signal. If you are above the compressor threshold, then the compressor turns down the volume again. This is a no-win situation (unless you have a make-up gain stage at the compressor's output). Even then, it is a better solution to set up the EQ after the compressor.
Click to expand...
Acorec is absolutely correct that EQ boosting before compression can and often will be adversely affected by compression. This is why you want to save EQ for sweetening the sound until after compression.

However, using EQ to cut *unwanted* non-peak frequencies is often (but not always, all these rules are general) more effectively done before compression; otherwise the compression will have a tendancy to relatively boost the unwanteds, making them that much more unpleasant and that much harder to take out.

G.
 
SouthSIDE Glen said:
Acorec is absolutely correct that EQ boosting before compression can and often will be adversely affected by compression. This is why you want to save EQ for sweetening the sound until after compression.

However, using EQ to cut *unwanted* non-peak frequencies is often (but not always, all these rules are general) more effectively done before compression; otherwise the compression will have a tendancy to relatively boost the unwanteds, making them that much more unpleasant and that much harder to take out.

G.
Click to expand...
This is exactly why I said earlier that Glen gave us 3 years worth of free advice - it's so subtle and easy to miss. EQ both before and after compression - but before only if you are 'fixing' something. The only 2 repairs I like to do right now before compression (always learning so this can change) are hum notching and gently adjusting the slope of the audio (wide bandwidths). In other words if the general slope is a little bass heavy (2db/octave) or treble heavy (5db/octave) I'll throw a shelf or 2 on there to get the slope I like before the compressor. Then put the 'sweetening' EQ on post compressor (as Glen calls it). Fix - Compress - Sweeten (just like a good cup of coffee) :D

For some of you guys with great recording skills there won't be a need for 'Fix'. I rebalance old tapes so by definition I'm always fixing...
 
Hmh.. I'll throw something into the soup too.

First of all the compressor is a very versatile and needed tool. It can also do alot(!) harm if used incorrectly. One big part of music are dynamics. Wrong compressor usage will kill these very effectively, so use your ears when playing with it.

Compressor is used to control the dynamics of the material. Very basically it turns the volume down and back up depengin from the parametres. You might even think it as a fader moving up and down very fast.

Compressor has usually at least these 4 basic parameters; treshold, attack, release and ratio:

- Treshold: This is the line where the compressor starts to work. Anything happening below this point and the compressor just sats there doing nothing actively (it may give the personal sound of the compressor used thou (eg some vintage "coloring")). So if you want to compress only some peaks you have in the material try not to dig too "deep" with this. On the other hand if you want to have more squashed and non dynamic sound dig deep enough.

- Attack (usually measured in ms): This is kinda the compressor's reaction time. When the compressor notices something going over the treshold you have set it starts to compress (turn down) the signal, but it only does it after this time has passed(attack). So if the attack is 50ms the compressor will wait 50ms after the signal has passed the treshold until compressiong it down.

- Release (again usually ms): This is similar to the attack. This means how long the compressor will compress the signal after the signal has gone below the treshold. So if this is 100ms the compressor will compress the signal 100ms after the signal has gone back under the treshold.

- Ratio (usually measured in 1:1, 2:1, 4:1 etc): This control means how much the compressor will lower the volume after the signal has passed the treshold. eg 2:1 means that if a signal passes the treshold point by 2db the compressor compresses (turns down) the volume by 1 db. So it is 2:1, meaning basically that the volume is cut in half (infact it isn't neccesarely this linear, but anyway heh) One good ratio settings to start with are 2:1-2.5:1 with things that need less and more subtle compressing and around 4:1 with other "normal" things (I propably shouldn't give such things since compressing is all about listening. So don't do it with eyes, do it with ears).

So putting it all together with an really basic example. Lets imagine our compressor being as a fader in a mixer with meters. First you sit there and decide that every time the needle (or whatever) hits -6db you start moving the fader (lowering the volume). This is our treshold. Okay now the drummer hit the snare and the needle jumped to -4db so quickly you start moving the fader down. Your reflexes had 50ms delay (man you're fast!). This is our attack time. It takes 50ms for you to start moving the fader. You move it down 1db. Now we can think that our ratio was 2:1. The peak went over the treshold by 2db and you lowered the volume by 1db. If it had gone over by 4 db you would have lowered the volume by 2db. So 2:1 stays. Got it? Good. Okay the release time here is the time it takes for you to move the fader back to original position after the signal has again went under the -6db. This was quite simplified example, but I think it may help to understand what it is basically all about.

Okay and finally some different things to use compressor:

- Try using quite fast attack time and moderate/hard ratio and treshold with single drumtracks (or any other intstruments you want to have more attack/snap or whatever) to give eg snare more punch. Again I shouldn't do this but... : Try 10-20ms, 4:1, treshold enough to achieve around 5 to more db gain reduction. Release time should be fast. If you stop and think what the compressor does here; it lowers the volume of the snare drum after that 10-15ms thus creating a "false dynamic". It makes those 15ms to be louder than the rest of that snare hit. Understand the idea?

- Try using around 2-2.5:1 ratio (or more if needed) and moderate attack time (don't kill all the attack here dudes!) and quite long release to smooth things up with the vox or instruments. Adjust the treshold as needed (again you just have to learn to listen, sorry m8s)

- If you just want to tame some of extereme peaks try using a high ratio (effectly turning the compressor into a limiter), fast attack and fast enough release to compress (limit) only the peaks.

- Generally longer release times tend to make the sounds smoother.

Hope someone did get something out of this :)
 
Last edited:
You must log in or register to reply here.
Back
Top