Clipping/distortion/lim/comp - what you do to your tunes

You know that 'C' that turns red when its clipping? Do you watch that or just go by ear to get louder?




[Title edited to reflect how the thread developed - Macc]

[Macc]

Errrr...., what are you on about?

Clipping is - unless done extremely carefully - bad. Red lights are bad!

If your tune ain't loud enough, then that's another issue...

:flashbacksofthebhimathreadcrusayay:

[Macc]

Yeah, where is that...? Much easier to point Profane there

Sorry Profane, I didn't mean to be rude, I just didn't quite understand your question.

[Macc]

Read this Mr Profane

http://www.subvertcentral.com/forum/view...hp?t=22064

Classique

timeless thread

clipping is bad!

on the master out bus the red c = clipping. ie your sound card is clipping. you are loosing sound.

0 is the limit. you cannot go above 0.

turn your master bus down so it never clips.

if your tunes is not "loud" enough. then good usage of compression and limiting will sort that out. you need to get as close to 0 as possible without going over it. bare in mind that too much excessive comp/limiting can have detriment al effects on your tune.

[Macc]

Do let us know if you need examples of that last part Profane... There's been a number of examples lately

Yeah, read my thread; some good info in there.

- unless done extremely carefully -

Can you elaborate on that? I would have thought that it's only not noticable as distortion if only done in tiny ammounts...? But doing it in tiny ammounts would, surely, defeat the point of doing it in the first place?

[Macc]

Clipping?

bhima?

I can feel a moan coming on

[Macc]

Anyway, so clipping can be used VERY carefully, if you want to try it.

AS ALWAYS, USE YOUR EARS YOU DUMB BASTIDS



Say you have a very short sharp spike in your audio, a re-mix is out of the question etc etc. For the sake of argument, at 44.1kHz the problem causing spike is 30 samples long. 30 consecutive clipped samples is a lot, right? I looks it in the wave editor... Well, at 44.1kHz, 30 samples works out as...

[1/44100] x 30 = 0.00068 seconds, or 0.68ms (or 680 microseconds )

That's not very long at all. Now whether or not this specific instance of clipping is audible depends on the signal, it may or may not be. So you have to use your ears. It depends on the frequency content of the signal (bass? hi hat?), how far over 0 it goes, the differential of it as it crosses 0 blah blah.... So see if you can hear it. If not, great.

However, say this isn't the only spike in your audio - there's quite a few of them. Chances are that although it is 'inaudible', the fatiguing effect of that skanky distortion (and yes it IS nasty, bhima ) will 'mount up' over time. So again, you have to use your ears. Whereas an isolated spike might stand a lot of clipping, the more you do it the more blantant - and straight up fucking annoying - it becomes.


'YEH BUT Y NOT YUZE DA L2 LIMITAH GHUY?'

Well here's the MAIN thing for me.

A look-ahead limiter (ie most plugin limiters) have no attack time, neither does clipping (when it hits zero, it hits zero). BUT limiting DOES have a release time - clipping doesn't.

As such, with our spike example above, the limiter will see it, think 'shit shit shit, big up spike bizzle a gwanin'. Then it scales down the signal at the onset of the peak, and continues to do so for time T, where (put simply) T = time(duration of peak) + time(release time).

So, in actual fact the limiter is acting on the signal after the peak has been gone. AND while it is acting it is scaling the entire signal, whereas clipping does nothing below the threshold, at any time, ever.



Reading this you'd think that clipping renders limiters useless then if it is so great. Well, not quite. By virtue of the scaling it doesn't generate the same nasty distortion as clipping, although the release time - when set too fast - can lead to the limiter releasing between the successive peaks of low frequencies and modulating the audio itself, not the gain. This is what you hear all over loads of fucking tunes on here, and the primary source of the sound I moan about incessantly.


The main point is that the (harmonic) distortion caused by limiters is less unpleasant than that caused by straight up clipping and therefore allows the person to do more gain reduction before the distortion becomes really annoying. There are numerous other artifacts caused by excessive limiting that do piss you off before that becomes an issue though.... Lack of dynamics, punch, too short a release with too much GR etc etc


Personally I fucking hate clipping anything, and swear I can hear it even when I probably can't - but that may well just be indoctrination. I know for sure that doing no more than 2dB of gain reduction (in the right way and not too much too often as described above), clipping can sound much more open than limiting.

Why does the fatiguing effect occur (even though the distortion it is inaudible)?


PS ...indoctrination....mmm....thats a big word i dont know the meaning of.

[Macc]

Why does the fatiguing effect occur (even though the distortion it is inaudible)?

Well it depends on how you define audible I suppose. By that I mean either saying 'I can definitely hear it' or 'detected by the ear whether you hear it or not'.

In this case it is the latter. The effects of straight digital clipping are perceived as ugly (and, well, are IMO), but without an extended period of clipping it is hard to say 'it's there, listen to it!'. However the ear still detects it, and it can get fatiguing.

I suppose the best analogy is someone reeeaaally gently poking you in the arm. It's fine at first but after ten minutes you want to smash their face in.


Of course it all depends on the signal, the degree of clipping the nature of it raa raa raa.


Just stick to using your limiters chaps, but watch yer release times and make sure you're not fucking up your tune

the release time needs to be long enough right?

that trick you mentioned where you apply ultimate walloping and set the release to the minimum you can without it distorting applies here n'est pas? or is it the other way round i forget what the actual trick was

[Macc]

That's for finding the 'magic' release time for a compressor, but they aren't entirely unrelated.

As with all things (well, generally) the release time needs to be

- short enough that longer term gain modulations aren't heard (so when the bass hits everything sucks out for a bit then fades back in)

- long enough that it doesn't treat each successive peak of low frequency signals as peaks in their own right.

That is, when the release time is shorter than the period of a given frequency (R < [1 / frequency] ), it will read the actual cycles of the wave as transients in their own right. Rather than modulate the gain of the audio signal, it will start to modulate the signal itself (think about that tyler, I know you'll get it). This is (largely) what gives you the distortion when compressing a lot with a short release.

Here's some rough calculations done out of my own interest (and cos I can't get on my music PC as Martsman is STILL afuckingsleep) ;

NOTE: DON'T TAKE THIS AS GOSPEL TRUTH, IT IS FOR GENERAL ILLUSTRATIVE PURPOSES ONLY!

We've got a compressor, and we're going to see what the shortest release time we can get away with - if we want to avoid these things - is.

Compressor on master output - 40 Hz is the lowest frequency in your tune.

1/40 = 0.025 seconds = 25 ms. If you release is shorter than that you run the risk of the compressor jumping about between the peaks of the bass.

Now say we have some 30Hz content;

1/30 = 0.0333r seconds = 33ms. A fairly big difference for the sake of 10Hz!


Say we are using it on a kick drum with prominent 100Hz content and very little below that;

1/100 = 0.01 seconds = 10 ms


What about a snare with big 300Hz content;

1/300 = 0.0033333r seconds = 3ms


So you can see from here that using very short releases - say for example 10ms release on the snare is fine so far as the 'audio signal modulation distortion' (catchy) goes. But when there is more low frequency content, you are obliged to use longer release times.



All this talk is meaningless really - when that martsman bastid wakes up I'll do a quick audio file and do it so you can hear wtf I am talking about.

for some reason i never made the connection between the period of the cycles and the compressor/limiter release before

its obvious why it doesnt want to be too long but i didnt notice that it could be short enough to fit in a single cycle

thinking about it though now i'm not sure about attack thats probably more to do with how fast the transient comes in/out? although so is the release

i managed to get some nice distortion out of the liquid mix last night :d probably just set the release too short sounds good though

Yeah, audio files would be nice. Thanks Macc./Martsman

[Macc]

CZHECKQUE IT

http://rapidshare.com/files/71945461/Re3...e.mp3.html

Guess who forgot to turn numlock off, hoho.

So look, this is a C0 sine wave being generated by the A1 synth, roughly 40 Hz I think it is.

It is put through Voxengo Marquis compressor - I did it with Waves RComp but the automation didn't work .

The compressor is set so that it is doing 10-11dB of gain reduction.

The file starts with the release at max - 8 seconds on Marquis. I turn the release down slowly and you can hear the expected distortion coming in, and it's no unpleasant on this source.


Then at about 36-38 seconds into the file, I get below 10 ms release. You hear the artifact I am on about at the top end. I then nudge the release back and forth a few times, going from no more than 15 ms, no less than 4ms. Listen to how quickly that effect fades in and out, even with such a small release time difference.


Again, listen to the high end. It doesn't sound like a sine wave any more, there's some raspy high end shit going on where the compressor is jumping up and down between the peaks of the waveform, modulating the audio itself. You can hear it rasping/drilling, not like a constant signal, or harmonic distortion.

The gain reduction meter also visibly jumps up and down when going below that point.

THIS IS WHAT YOU GET ALL OVER YOUR PRECIOUS TUNE WHEN YOU LIMIT BY 6dB WITH HEAVY BASS CONTENT AND TOO SHORT A RELEASE TIME.

CZHECKQUE

can hear it really clearly in that, sounds kind of nasty too

THX

Check it when i get home from uni.

wow thanks for explaining that one macc!



i just don't get it why people do use such rediculous small releasetimes

again, listen to the high end. it doesn't sound like a sine wave any more, there's some raspy high end shit going on where the compressor is jumping up and down between the peaks of the waveform, modulating the audio itself. you can hear it rasping/drilling, not like a constant signal, or harmonic distortion.

neat example ,macc should be a teacher

...
Here's some rough calculations done out of my own interest (and cos I can't get on my music PC as Martsman is STILL afuckingsleep) ;

...

All this talk is meaningless really - when that martsman bastid wakes up I'll do a quick audio file and do it so you can hear wtf I am talking about.

i knew i was a nuisance to you

[Macc]

- short enough that longer term gain modulations aren't heard (so when the bass hits everything sucks out for a bit then fades back in)

French House creux.