Skip to main content

So I have been trying out every thing I learn here, but have a question which might be a little basic.
Feel a bit confused about the idea of using a compressor. Should I let the peaks (say drums) be as they are but without clipping and bring up the other instruments with a compressor, or let the instruments be as they are and bring down the peaks with a compressor? Should the bottom be pulled up or the top down? In analog I understand one will not want to pull the bottom up because of the noise floor. Do we use compressors the same way or are there other methods that can be tried since there is no noise floor?

Thanks!

Topic Tags

Comments

pcrecord Sun, 06/05/2016 - 14:01

Hi JJitter,
A compressor like other tools should not be used unless there is something to be fixed.
Be carefull, you can remove the life out of a track or a mix very easy with those.

Compressors can do many things and you need to master the relation between the settings.
It's all about at what detected level (thresold), at what time the compressor starts to work (attack) and how much it lets through (ratio) and when it stops to act (release).

If you want a puchier track, you'd use a 30ms to 100ms attack and a longer release (timed with the speed of the content)
If you want to control only the peaks (transients) you can go faster attack and faster release.
If you want to glue things together, you'd use a slow attack etc...
Most of the time, I use light compression that will remove 2 or 3 db of the content.

BUT !! It's not a volume controler for softer parts of the song, use volume automation instead.

The best thing you could do is to practice, practice and practice...

If you want ask more specific questions, name the instrument and say what you think is wrong. Post an exemple even, it'll be fun ;)

kmetal Sun, 06/05/2016 - 19:26

Hey jitter. Compression is something that you kinda get a feel for, as opposed to hearing it.

Something I learned from a guy I assisted for a while, was his compression technique, which I use sometimes.

Seemingly no matter which compresser he's using he sets the attack to the slowest, and the release to the fastest. Then using the threshold, he lowers it until he's getting the desired amount of compression, usually somewhere between 1-6db of GR. This effectively allows the all important transients through the mix, sorta rounding things out a bit.

Compression is tough becasue if your hearing it, it's generally 'too much', so it's very much a vibe thing, which personally took me well over 5 years to start to really 'hear'. In 100% truth, I often don't hear the 2-3 db of compression I'm adding to things, but it is subtly there. Compression is more about how your making the instrument sit with everything else, versus what your doing to the instrument itself.

Modern music is quite compressed. So with regard to your specific question, you'll want everything to meet somewhere in the middle. The louder stuff can be knocked down a bit, the lower stuff can be brought up.

The tip I learned from the mix e engineers handbook, (by Bobby owinski) a must read for any engineer, is to hammer the threshold on the compresser to like a lot of GR, like 12-18db, then adjust your attack and release, this is going to allow you to 'hear' the compression. Once you get the instrument breathing, or sustaining, or more consistent, whatever your trying to accomplish, once you get your attack and release doing what you want, you dial back the threshold, so you've got a more practical amount of GR.

One thing to keep in mind, is compression, is very much divide and conquer, and in general your better off using two or more gentle compressors in a row, than one compressor doing tons of heavy compression. Kinda like how piant is best applied in thin coats, than all at once.

Again, there's no rules. But there are certain standards, and it's good to be aware of them while your completely ignoring them :)

kmetal Sun, 06/05/2016 - 21:18

P.s regarding drums, another excellent tip from the book, which I use all the time, is when compressing something like a snare or kick. Set your release, so the GR meter returns completely back to 0 ie no compression, by the time the next snare or kick hit starts. This helps you make sure your not mushing the transients.

DonnyThompson Mon, 06/06/2016 - 02:33

I would strongly suggest that you check out the links that Sean posted; in particular, the SOS link, ( SOS has a well-deserved industry reputation for being a source of and providing accurate info) from which I've pulled this excerpt, an approach I think is worth considering:

"concentrate on the balance of the tracks in your mix. If the tracks balance fine as they are, no one will arrest you for leaving them alone! The trick is to wait until you spot a fader that you can't really find a suitable level for (the sound may disappear in some places, or have sections that feel too loud): that's where you may need to compress. In the first instance, though, see if you can solve the level problem by splitting the audio onto two different tracks and balancing them separately. This is a common technique often referred to as 'multing'. It's easily done in most DAWs, and can head off a lot of rookie compression mistakes. Again, you may find that you don't need any compression at all to find a balance that works..."
( Donny's own insert here - this is also where you could work with volume automation using your DAW's editing features).

I think that one of the most frequent beginner/novice mixing mistakes is to reach for compression too quickly, which can add all sorts of issues when you factor in that those same beginners who are reaching for it in this knee-jerk fashion don't truly understand how gain reduction works, or what the settings do, and who use compression expecting an audible effect.
Some gain reduction models can add character, and are used because they add a certain sonic vibe... but you need to know the effect you are going for before you add it with that goal in mind. A good portion of the time, sufficient and effective compression isn't really obviously audible, but serves its purpose in a transparent way, reducing gain without there being a noticeable effect.

My personal recommendation would be for you to use more transparent "non-character" compression processing - at least for right now anyway; avoiding those compressor and limiter plugs that are modeled after well-known hardware models... plugs that model FET, Tube or Opto based compressors ( 1176's, Fairchilds, LA2A's, etc.) and focus on learning and becoming familiar with "basic" compression plugs - at least until you get used to how gain reduction works, and how it effects your signal...

But that's just a personal observation.

Sean G Mon, 06/06/2016 - 03:00

DonnyThompson, post: 438867, member: 46114 wrote: My personal recommendation would be for you to use more transparent "non-character" compression processing - at least for right now anyway; avoiding those compressor and limiter plugs that are modeled after well-known hardware models... plugs that model FET, Tube or Opto based compressors ( 1176's, Fairchilds, LA2A's, etc.) and focus on learning and becoming familiar with "basic" compression plugs - at least until you get used to how gain reduction works, and how it effects your signal...

Donny makes a good point...my advice would be play around with the native compressor in your DAW to get a feel for its function.

Once you get the hang of it, how it applies and how you can use it to benefit your tracks, you can then experiment with a few different model comp plugs down the track.

My native compressor plug-in in Studio One can can accomplish what many top-dollar third party compressor plug-ins can.

kmetal Mon, 06/06/2016 - 04:17

DonnyThompson, post: 438867, member: 46114 wrote: I would strongly suggest that you check out the links that Sean posted; in particular, the SOS link, ( SOS has a well-deserved industry reputation for being a source of and providing accurate info) from which I've pulled this excerpt, an approach I think is worth considering:

"concentrate on the balance of the tracks in your mix. If the tracks balance fine as they are, no one will arrest you for leaving them alone! The trick is to wait until you spot a fader that you can't really find a suitable level for (the sound may disappear in some places, or have sections that feel too loud): that's where you may need to compress. In the first instance, though, see if you can solve the level problem by splitting the audio onto two different tracks and balancing them separately. This is a common technique often referred to as 'multing'. It's easily done in most DAWs, and can head off a lot of rookie compression mistakes. Again, you may find that you don't need any compression at all to find a balance that works..."
( Donny's own insert here - this is also where you could work with volume automation using your DAW's editing features).

I think that one of the most frequent beginner/novice mixing mistakes is to reach for compression too quickly, which can add all sorts of issues when you factor in that those same beginners who are reaching for it in this knee-jerk fashion don't truly understand how gain reduction works, or what the settings do, and who use compression expecting an audible effect.
Some gain reduction models can add character, and are used because they add a certain sonic vibe... but you need to know the effect you are going for before you add it with that goal in mind. A good portion of the time, sufficient and effective compression isn't really obviously audible, but serves its purpose in a transparent way, reducing gain without there being a noticeable effect.

My personal recommendation would be for you to use more transparent "non-character" compression processing - at least for right now anyway; avoiding those compressor and limiter plugs that are modeled after well-known hardware models... plugs that model FET, Tube or Opto based compressors ( 1176's, Fairchilds, LA2A's, etc.) and focus on learning and becoming familiar with "basic" compression plugs - at least until you get used to how gain reduction works, and how it effects your signal...

But that's just a personal observation.

Well said D

+1

jjitter Mon, 06/06/2016 - 04:38

thanks @Sean G will check out the articles.
Wasn't really relating a compressor with tempo. Thats a good starting point to tweak the attack and release knobs.
Yeah just been trying out the stock processors. But everytime the processing becomes too less or too much.
Need to spend a lot more time to come close to a good setting.
pcrecord will try posting an example

jjitter Mon, 06/06/2016 - 06:19

My big worry is the drum mix. The problem is it doesn't sound glued or 'one'. Here is what I try to do - Eq, level and pan individual mono drum sounds, put all drums on one mono bus and kick and snare on the other so i can aux send it to the sidechain comp on the bass guitar track if needed. Again sum these two mono buses to a stereo bus and try to eq and compress this main drum bus. The DR seems to be fine. So it probably doesn't need compression for gain reduction to control peaks. but for gluing. Which took me to compression. I tried a lot of settings on some of the compressors I have. even tried the presets for drum bus. But still no luck. Something to do with the input gain on the compressor? Or the busing?

Sean G Mon, 06/06/2016 - 06:21

jjitter there is a good thread here http://recording.org/threads/parallel-compression.59296/ started by DonnyThompson a little while ago that discusses parallel compression and its different applications. Parallel compression can be used to enhance tracks and is a great technique for drum buses. If you are having issues getting a drum track to sit right you could try using parallel compression and blend it back in, therefore giving you fuller sounding drums while allowing you to attenuate the drum bus given that you are effectively doubling the sound.

Its a great technique that is not only effective on drums but also vocals. You can use it on any track really.

Sean G Mon, 06/06/2016 - 06:39

jjitter, post: 438872, member: 49846 wrote: Eq, level and pan individual mono drum sounds

You really want to keep the drum tracks and the bass track for that matter centred. Maybe a little of the high hat to the left or right by say 10 and you can do your toms say 10L / centre / 10R and overheads to the L & R, but the drum track as a rule of thumb should be at the centre of the sound stage, with other instruments to the left and right of the drum & bass.

jjitter, post: 438872, member: 49846 wrote: Again sum these two mono buses to a stereo bus and try to eq and compress this main drum bus.

You really don't want to be using EQ & compression just for the sake of it. When you compress drum tracks they can become thin and set back in your mix if you are not too careful.

jjitter, post: 438872, member: 49846 wrote: I tried a lot of settings on some of the compressors I have. even tried the presets for drum bus.

Presets can give you a starting point, which is good if you are starting out, but don't rely on them as a fix. You still may need to dial things in to some degree.

What ever that preset was designed around may be entirely different to how your track sounds or sits in the mix due to many different factors. So while they are a guide to a degree, try not to rely on them. If you are relying on presets then you are limiting yourself to just that number of half a dozen presets...dialing in settings yourself will give you endless possibilities.

By getting yourself used to relying on your ears and not your presets you will train yourself to listen to what is happening to your track in the mix, and the overall effect the plug-in is having on it.

Once you are confident you may just find yourself creating and saving your own presets that you can use later on as templates for different instruments and applications.

The more you practice playing around with your plug-ins, the more confident you will become... trust your ears not your presets.

kmetal Mon, 06/06/2016 - 11:16

You really want to be sure how your Daw handles pluggin latency, particularly reguarding parellell compression. If deley compensation is not applied properly you could have some nasty phase artifacts. I incur this when I have to use DP 7.

If for instance you leave your snare bus dry, and send it via and aux bus for parellell processing, there's got to be a pre-emptied delay on the dry track, because it takes time for the computer to process the effects, and combine them back with the original. Since the dry is a "straight shot" it would arrive at your ears(mix bus) sooner, than the processed signal, cause passey, chorusy type deal, which is detrimental to the solidity of the recording.

Also, you'll want all your drums on a stereo bus to preserve your stereo panning. Additional mono buses for kick and snare are fine because they're generall center panned.

pcrecord Mon, 06/06/2016 - 12:45

jjitter, post: 438872, member: 49846 wrote: My big worry is the drum mix. The problem is it doesn't sound glued or 'one'.

A good part of the drum sound comes from how it's been played, mic techniques and room.
I often use a ribbon as a mono room mic ; well placed it gives some body to the bassdrum and snare without grabbing too much cymbals. It's my first step to glueing things together.
I often compress this mono track a bit more and just place that sound Under the whole kit..

jjitter Tue, 06/07/2016 - 04:13

pcrecord actually the only thing i can record with a mic is the guitar, keyboard amp. Dont have a drum room yet. So the drum parts come from vst.

So I started eq'ing a drum section to get the kit 'closer' or sounding 'one', thinking that I will take care of the peaks after getting the good sounding kit. Got the kit sounding right. Now since I was only using substractive eq, there was gain reduction and the peaks were gone! Got rid of the peaks on the way to getting the kit sounding right. Two Birds! Guess most of the frequencies that were causing the signal to peak were not even needed.
DonnyThompson I see what you mean by reaching the compressors too soon.

pcrecord Tue, 06/07/2016 - 18:27

There is nothing wrong with having peaks ! ;) It's in the nature of drums...
The first goal of mixing is to balance the tracks so we hear all the instruments in a musical way.
Fear of peaks often comes to persons pursuing higher volumes and falling for the loudness war.

At this point I thinks you should post a drum track alone and that same track in the context of the song so we can hear your challenge.

Sean G Tue, 06/07/2016 - 20:30

pcrecord, post: 438934, member: 46460 wrote: At this point I thinks you should post a drum track alone and that same track in the context of the song so we can hear your challenge.

I agree with my esteemed collegue from Quebec, maybe if you post your drum track in question we may be able to offer some further creative assistance to guide you along the path of percussive enlightenment...:D