How to notice the effect of a compressor?

Performance & Techniques Music Production
, , ,
1

How to notice the effect of a compressor?

Hey guys,

I know compression has been spoken about A LOT! though im still unclear.

Okay i know what all the controls on a compressor are for. For eg i know what a ratio/knee etc does. i know know a compressor reduces the difference between the highest and the lowest peak of a signal so the volume can be raised. i have read up a lot..

Though what i cant get is, i can’t seem to hear the effect of a compressor much. now when i say that i mean, when i put on a compressor i dont really know when its effect is too much or too little. Obv, when the make up gain is added it gets louder and when not it get softer. Thats very obvious. Though i know what attack release and all the parameters do, i dont know how to customise them for diff signals. i own the KRK rp8s though cant treat my room (not an option sadly). My technique of compressing is: pull threshold completely down, max attack, min release and then slow start reducing attack and suit everything to need.

Whats your technique and what would you advice me?

2

I have some audio examples that might help in my guide here:

http://tarekith.com/assets/pdfs/DynamicsControl.pdf

3

One obvious tip, you probably know is to make sure you have the meter set to show your gain reduction.

I’m going to out on a limb here and say that sometimes not hearing the effects of compression can be a good thing. If you’re seeing the GR meter bump down a few dbs and you’re not really hearing any effect on the sound, you’ve got some nice “transparent” compression - meaning you’ve leveled the sound without any noticeable side effects and you’ve given other elements in the mix more room.

As far as technique I do pretty much the same as you. I start with a slow attack (especially with drums) so I don’t squash the transients, and a 1:2 ratio and listen to the dynamics. I adjust attack if I need more or less transients, release if I want the compression to sit better with the rhythm of the track. Ratio up or down depending on how mush dynamics I want preserve on the original signal. I start with a pretty aggressive threshold so I can hear the compression in a very dramatic way and then adjust it to be more subtle when I’ve got the other settings more or less right, then repeat until I get what I want.

4

That’s sort of the point of “good” compression, you’re not supposed to hear it.

You and Nick V have the basics down, but I want to point out an important aspect. First off, I listen. Once you get accustomed enough to mixing you get a grasp (a feel) for how you want to apply compression when needed. There’s no real proper formula to compression as the “why” and “how” change with every application.

That being said, there’s obviously some methodology to it, which you sort of figure out for yourself. I personally start with the attack and release before even engaging the compressor. I do this because if I turn it on, roll down the threshold and the attack/release are in wonky pre-settings or set to the absolute fastest time, it’s going to sound unnatural right off the bat, which is like starting with a messy canvas (IMO).
So I set the attack/release in a neutral area (usually somewhere between 2 and 4 o’clock, 5 being fastest time), turn on the compressor and slowly roll down the threshold. At this point I can hear the compressor working, and I’ll set the threshold a a little lower than I’m planning on using. By now I probably hear what I like/don’t like, and will adjust the attack and release accordingly.

Attack/Release settings require the most thought. Going back to my first point, based on what you heard and what you’re looking to achieve you only need to translate that interest into the right settings. General examples would be:

  • Transients are out of control: Fast attack/Fast Release
  • Resonance/Energy needs taming (vocals for instance): Slower attack/Fast-ish Release
  • Needs more punch: Fast-ish attack(letting first transients through)/Medium Release
    (Again just very crude general examples, don’t use these as guides)

Once I have the settings close to what I want I roll the threshold back up to a more reasonable point. As a rule of thumb I try to never start compressing over 4dB’s gain reduction. Then I perform another crucial step: listening again. Dial in the make-up to where the level is the same as pre-compression and turn the compressor on/off (bypass). Usually I can clearly hear what changes I’ve made and then I just ask myself: “Is this what I wanted?” No, keep making adjustments, Yes, done. When I work like this I generally just fine-tune a little and end up rolling back the threshold more usually. If you really check to hear what’s going on you’d be amazed by how much a compressor is doing, even when used slightly. Which is why you always see rookies compress the hell out everything.

Obviously you can use a compressor as an effect as well, but that’s completely different. Usually you’re trying to overdrive the circuits which distorts the track and can add drastic pumping/ducking depending on attack/release settings. (Though you’ll get that either way if you’re using improper settings).

All in all, compression is tricky, as you really need to focus and train your eyes to understand what’s happening. But as you keep doing it you’ll get a feel for it. Just remember to always listen and reference check, 'cause that’s really the only way to learn and retain your experience.

5

Interesting! Thank you for explaining your process in good detail. This will definitely help and ill consider your points while compressing the next time. :slight_smile:

6

For “bad compression” this contains one of the more “obvious” examples of the effect that I have found. Listen at “moderate to high” volume levels on reasonable headphones or speakers…

7

Okay guys, thanks for all the help and tips. I just completed the mixdown of my track and all your advice has helped me a lot. I have a headroom of 5.56 db left. Would you please give it a listen and let me know how the mix is? I haven’t yet mastered it so its low in volume. I believe this is my best mix until now but i’m not a pro yet. This is my 14th track. Thank you. :slight_smile:

https://soundcloud.com/sagar4848/code-x-unmastered

8

I thought the high hats, crash, and that squeeky noise were a touch bright. Low end is a touch light, you could beef that up some. Overall it’s got instrument to instrument balance and lots of good definition, but not much warmth, like all the lower mids were high-passed out. A little less highs and little more lows should sort it out.

9

Haha yes i agree. As a matter of fact i too, was contemplating the hats…and about the low end too, but in the end of it i just got too frustrated trying to balance it all out and settled for this. I tried hard to get in some low mids but i guess i just had some missing elements there. I generally use the vintage warmer 2 (presets), camel phat or klanghelm sdrr for some warmth to a signal. Any other cool suggestions?

Thanks for the help. :slight_smile:

10

My favorite tool for boosting low mids is the Pultec. I use the waves version but I’ve heard the UAD version and it’s even better.

11

oh thanks! I shall check out Pultec! :slight_smile:

12

You’re track is unavailable, so I couldn’t listen. In any case, don’t shy away from getting closer to 0dB, even before mastering. Having headroom is great, but when you’re compressing your track (which I’m assuming you are), then there’s no reason you can’t have it under control and inch it a bit closer to the ultimate goal anyway.

My cool suggestion is to go back and work on individual elements before doing any stereo bus EQ-ing/Compression. Listen, listen, and listen again until you know exactly what isn’t gelling well. Even with the most basic DAW tools you can achieve good results if you simply apply them correctly. It’s also often a matter of doing less, than doing more.

I’ve always considered the bus compression/EQ-ing sort of like putting a cherry on the cake. It’ll make a good cake all that little bit better and more presentable, but can’t mask the flaws in a crummy cake.

Despite the Pultec EQ (which is a great suggestion Nick V!) being a fantastic EQ (if you use attenuation cough), I dare you to work with general, non-specialized tools to get your mix sounding exactly the way you want it. I dare you, because you learn best through limiting yourself, and still making it work. Because what kind of point do you make when you have access and use amazing gear (we’re really spoiled with plugins) if the same (if not better) mix can be achieved with stock DAW plugins? My 2 cents.

13

Well said.

14

Hey thanks for your response.

So when you say don’t shy away from getting closer to 0db what do you mean? I mean to say, i’ve always read everywhere that you should like have a general headroom of 6db before mastering and then let the limiters/compressors etc come into action. Thats how i i generally mix my track, i keep the aim of a 6db headroom(not my ultimate goal but just kept in my mind). Because if i would cross 6db i wouldn’t know how much is enough and where to stop and leave how much headroom for mastering. My general aim is balance all the elements, make sure no frequencies are clashing and its sounding nice and cohesive but at a low volume. If I’m wrong or should add some things please correct me.

Maybe it’s just the lack of experience talking on my part but i’m curious and getting my basics right. :slight_smile:

Yes i completely agree on making each element as strong as possible. Now i focus on that every time. And yes indeed we are really spoilt with plugins. :stuck_out_tongue:

Thanks for you help.

Btw yeah sorry pulled it down and re-uploaded the track here:

https://soundcloud.com/sagar4848/code-x

15

Nothing wrong with aiming for -6dBFS, it’s what I generally recommend too. But it’s not a hard and fast rule either, if you’re absolutely sure you’re not clipping anywhere in the song, there’s really no reason you couldn’t mix to -0.1dBFS to be honest.

Most people recommend -6dBFS as a guide simply because it’s a simple way to add a bit of a safety net when it comes to headroom. If you’re generally around that point with all your mix elements playing, then you can just focus on doing the mix and not have to pay constant attention to your master meter.

Other than expending a calorie reaching for the volume knob to turn your studio monitors up a bit because you’re mixing with a goal quieter than the usual 0dB, there’s really no downside to leaving a few dB’s headroom when you’re working at 24bit or above. At the same time, if you’re bus compressing and it the master level creeps above -6dBFS, like R01 said, don’t stress it too much if you’re not clipping anyway.

16

Tarekith explained it well, so I have no need to repeat the same stuff.

I personally don’t stick to -6dB headroom (or even less for that matter) when it makes sense. You can always count on the mix peaks hitting 0dB after mastering, so I choose to maximize signal/noise when possible.

Context is super important, which is a shame because it’s rarely stressed next to generally basic techniques. That being said, you won’t catch me compressing a classical recording, and I probably won’t even care if there’s a difference of 10-20dB in dynamics. But we’re talking about electronic music here, so obviously compression is needed, and I’d do everything I could to push the volume without killing the dynamics. Complete opposite aproaches if you ask me, but it works based on context.

I could outline “mixing through a bus comp”, (which is also context dependent) but that’s probably beyond the scope here. The only other crucial tip I can give besides listening is reference. The biggest favor you can do for your ears is pulling a similar song, import it into your DAW, set the volume to the same as your mix and A/B-ing. Instead of thinking about what to adjust based on your own reference, it’ll hit your ears like a smack in the face. It’s also an important part of developing your own reference, so don’t underestimate it.

Not in a spot where I can listen easily, but I’ll throw your track on when I get home.

17

Alright guys, thanks for you inputs. They’ve been helpful and i’ll keep these pointers in mind the next time. :slight_smile:

18

This really doesn’t make sense in the digital domain…

Even 16 bit gives you 96 dB of dynamic range, 24 bit give you 144 dB, and even the calculator Bill Gates uses to count his money wouldn’t be able to show you the dynamic range the 32 bit floating point algorithm of your DAW has to offer.

So even with 16 bit, leaving 6 dB headroom would mean reducing your dynamic range to 90 dB, which is way more than even classical music has, let alone electronic music that usually has a dynamic range of less than 20 or even 15 dB even before mastering…

Plus, where analog gear and microphones actually produce noise, the noise floor of productions that are created strictly “in the box” is a rather threoretical thing…

So basically mixing closer to 0 dBFS instead of -6 dBFS to maximize signal/noise in a DAW with digitally created instruments in 24 bit is like putting on a rain coat in Death Valley to make sure you don’t get wet… :wink:

+1 on this.