Mastering Without Clipping Converters

[sanddigger1]

Originally Posted by Lupo

Red hot chili peppers, dani california (single), track 1: OVERLOAD in regular metering(set to indicate this at 1 sample over), +0.8 and +0.9 in oversampled metering

how is it possible ?

I recently found a CD with approx +0.8dB peak level (and not occasionnally). (I need to digitally reduce it by -0.8 to have it stop overloading)

I do not understand how something can go over the last available bit. Can somebody explain ?

[living sounds]

The intersample peaks result from the waveform reconstructed of the samples. If you can, try out an audio editor like Audition that shows the reconstructed waveforms. It will all become clear.

[sanddigger1]

I think I've understood the intersample clipping. But I thought that normal metering taking place before the D/AC couldnt go over 0.

[MAzevedo]

Originally Posted by sanddigger1

I think I've understood the intersample clipping. But I thought that normal metering taking place before the D/AC couldnt go over 0.

This is true, which is why you have to be careful. Your meters can show no overs, yet you still might hear distortion once it goes back to analog.

[sanddigger1]

so how can normal metering overload ?

[MAzevedo]

What exactly do you mean by "normal metering"? If you mean digital, non-oversampling metering, then it will call some number of consecutive full-scale samples (usually 1 or 2) "over". If you clip, and then after clipping drop the level 0.1dB, the you will have no full-scale samples (but lots of consecutive samples at -0.1dB), and you won't see any overs on the meters. However, if the analog stage after the D/A doesn't have the headroom to deal with the amplitude of the reconstructed waveform (which will have peaks greater than full-scale), you could still hear distortion on playback.

[sanddigger1]

In fact, Lupo called it 'regular metering'. I mean metering before the DAC, previous any wave reconstruction.
What I was trying to say is that I 've got a CD which overloads, without more audible distortion than the one its squashed modern low end mastering produces. If I reduce the output level of the player, it still overloads as long as I dont reduce the level by 0.8 dB.
That's what I dont understand.
With every other CD which hits 0 FS, reducing by 0.1 dB is enough to avoid hitting 0 (as you said, and as I was expecting), but not with this particular one.


EDIT : I extracted the audio : the peak level is 0FS, and playback in the editor was absolutely normal (reducing by 0.02 dB was enough to stop hitting 0). I should have started by doing this test. It is a problem with the player, which happens only with this CD (maybe because it is heavily clipped ?). So, no discovery of a 'over full scale coding'. Thanks to anyone who has tried to help me. This intersample clipping thread is very useful and interesting.

[nickynicknick]

Originally Posted by DSD_Mastering

Besides looking at the meters, I thought you were supposed to master with your ears?

Regards,
Bruce

funny. I guess meters are so irrelevant that they shouldn't exist?. Especially in mastering suites. Really...what kinda shit is that?

[nickynicknick]

Wow a lot of replies...kinda moved off topic a bit, but still Ok. Thanks for all your input. So in the end (if this is) most of you clip converters. In fact, it seems to be the way you master records these days. Very interesting. I do not recall anyone saying outright that they don't clip converters.

Nick

[Paul Frindle]

Originally Posted by nickynicknick

Wow * lot of replies...kind* moved off **pic * bit, but still Ok. Th*nks for *ll your input. So in the end (if this is) most of you clip converters. In f*ct, it seems ** be the w*y you m*ster records these d*ys. Very interesting. I do not rec*ll *nyone s*ying outright th*t they don't clip converters.

Nick

There's been lots of discussions about this and I have been involved in a good few. From experience, most modern productions cause intersample peaks that may cause overload on playback. and the sound of that overload will be different depending on the kit playing it back. The listener can improve things slightly if they have CD/DVD player with an internal digital volume control you can wind down, but really it's a nightmare IMHO.

[Paul Frindle]

Originally Posted by SafeandSound

I don't or plan to clip convertors in,out, plugs or the output meter of a DAW.

Some music benefits from a average level increase, some doesn't.

We can ask why do we have to clip audio?, then you can choose to do it or not. It's extreme increases which have the worst side effects in terms of distortion and loss of transients. Extreme anything is rarely any good for very long.

Yes - clipping can actually sound great and was particularly used (in the days of analogue) on drum attacks and such to add some harmonic 'punch' to the transients in an otherwise soft sounding track. The thing is that in digital you can only get away with this if the level is reduced afterwards to prevent odd and sometimes unpredictable things happening down line. Essentially clipping in the digital domain is not really legal.

[zboy2854]

Originally Posted by Paul Frindle

Essentially clipping in the digital domain is not really legal.

Perhaps we should have a loudness "prison" for the worst offenders...do not pass Go, do not collect your mastering fees...

[Paul Frindle]

Originally Posted by zboy2854

Perhaps we should have a loudness "prison" for the worst offenders...do not pass Go, do not collect your mastering fees...

LOL - the thing is that you can have the sound of clipped instruments, but they cannot be at full level in the mix. If you wind them down to -3dB or less then normally they will be preserved when played out.

[sonicdefault]

Originally Posted by Paul Frindle

Yes - clipping can actually sound great and was particularly used (in the days of analogue) on drum attacks and such to add some harmonic 'punch' to the transients in an otherwise soft sounding track. The thing is that in digital you can only get away with this if the level is reduced afterwards to prevent odd and sometimes unpredictable things happening down line. Essentially clipping in the digital domain is not really legal.

Hi Paul!

I was wondering when you would show up! (i saw your earlier post too) Congrats on the new company! Hint AU Hint


-SD

[JB872]

Are there any plugins now that sound as good or better than clipping an AD converter in 2010? Or is clipping converters still a common practice in mastering?

[dave-G]

Originally Posted by JB872

Are there any plugins now that sound as good or better than clipping an AD converter in 2010? Or is clipping converters still a common practice in mastering?

Yes


@ both questions


Well ... kinda.


Though seriously .... it depends on the converters/plugin and what you're hitting it with. Some converters really do this better than they should be able to.

-dave

[SoundWeavers]

Originally Posted by gainstages

i've never done it. i was actually shocked to learn that this was even being done in mastering - I had heard some people deliberately doing it during tracking or mixing but is just seemed so wrong to do it in mastering - I was really surprised to find out its pretty common.

+1. I just never bought into the idea of clipping converters.

[JB872]

Originally Posted by dave-G

Yes


@ both questions


Well ... kinda.


Though seriously .... it depends on the converters/plugin and what you're hitting it with. Some converters really do this better than they should be able to.

-dave

Hi Dave

What plugins would you recommend trying? And how does hard clipping the 2 buss compare to using a soft clipper?

[Ben F]

It really is source dependant, this comes with experience. Generally music with midrange distortion does not respond well to clipping, and music that has already been clipped in the mix stage.

Just mess around with the Ozone 4 limiter- the slider goes from clipping to soft limiting.

[dave-G]

Originally Posted by Ben F

Just mess around with the Ozone 4 limiter- the slider goes from clipping to soft limiting.

@JB872 I'd sincerely second this recommendation. Ozone 4's limiter is pretty flexible in this regard.

-dave

[JB872]

Originally Posted by dave-G

@JB872 I'd sincerely second this recommendation. Ozone 4's limiter is pretty flexible in this regard.

-dave

Thx Dave. I'll try this out as well. I'm also considering trying to clipping my FF800 AD. Just wondering what the best way to do this ITB would be. So say my mastering chain looks like this ITB.

EQ>compressor>External Insert (FF800 AD)>Brickwall Limiter Set to 0.3

What would I use to increase the gain into the converter? The output of the compressor? Or the input fader in RME total mix?

[bass man]

Originally Posted by JB872

Thx Dave. I'll try this out as well. I'm also considering trying to clipping my FF800 AD. Just wondering what the best way to do this ITB would be. So say my mastering chain looks like this ITB.

EQ>compressor>External Insert (FF800 AD)>Brickwall Limiter Set to 0.3

What would I use to increase the gain into the converter? The output of the compressor? Or the input fader in RME total mix?

I recommend using hi end hardware line driver for this or the make up gain on a good hardware comp.
I use my studer 962 ,or neve portico 5042 line driver etc for that extra 1 to 3 db`s into my apogee rosetta with soft limit on - sounds great most of the time.

[lucey]

(not a fan of the Ozone)

[dave-G]

Originally Posted by lucey

(not a fan of the Ozone)

I've had it less than a year, and only use it for the limiter. But I usually like it.

To JB872, I'd only add a prediction: if limiting or ADC clipping is what you're looking to as your primary means of loudness, you might be disappointed. It's generally most effective for the final dB or two. Definitely *not* a magic bullet.

Good luck!
-Dave

[lucey]

Originally Posted by dave-G

I've had it less than a year, and only use it for the limiter. But I usually like it.

thumbsup

[pmx]

maybe it's me being weird, but i NEVER clip my a/d. it sounds utterly crap, and mine (a hedd) is even said to be one of the better ones at it.

there are way better means to achieve (more) loudness, imho.

[Cellotron]

Originally Posted by pmx

maybe it's me being weird, but i NEVER clip my a/d. it sounds utterly crap, and mine (a hedd) is even said to be one of the better ones at it.

Actually I've heard a couple people say that the HEDD doesn't clip all that well in comparison to some other ADC's - but not having heard one being put into clipping in person this is just hear say. So far the ADC's that to my ears leave the least amount of undesirable artifacts in my testing have been the Burl B2 and the Lavry Gold (I had an earlier dB Tech model for a couple weeks before selling it for some needed cash). The Burl's coloration on the mid/upper mids (it's more saturated but softer transients in this area) made me pass on it - while the Lavry Gold's price (and some reported reliability issues) made me pass on it as well.

Best regards,
Steve Berson

[Table Of Tone]

Originally Posted by Cellotron

Actually I've heard a couple people say that the HEDD doesn't clip all that well in comparison to some other ADC's - but not having heard one being put into clipping in person this is just hear say. So far the ADC's that to my ears leave the least amount of undesirable artifacts in my testing have been the Burl B2 and the Lavry Gold (I had an earlier dB Tech model for a couple weeks before selling it for some needed cash). The Burl's coloration on the mid/upper mids (it's more saturated but softer transients in this area) made me pass on it - while the Lavry Gold's price (and some reported reliability issues) made me pass on it as well.

Best regards,
Steve Berson

I'm an AD 122 user myself and nothing else I've tried, touches em, sonically!
The AD 122 seems to retain more of the natural dynamics of the mix and more of any movements I've made in the analogue domain.
I use a slightly different version to the standard one.
They are expensive and a little temperamental.
I'm probably gonna have to get another spare one!

[Paul Frindle]

Originally Posted by Table Of Tone

I'm an AD 122 user myself and nothing else I've tried, touches em, sonically!
The AD 122 seems to retain more of the natural dynamics of the mix and more of any movements I've made in the analogue domain.
I use a slightly different version to the standard one.
They are expensive and a little temperamental.
I'm probably gonna have to get another spare one!

I can't believe you guys are actually clipping A/D converters for sound effect!? And even going a far as rating them on the basis of how they respond to overload errors?

This is a very bad idea IMVHO - and what you will get will be a lottery involving what the converter actually spits out and what will happen when the erroneous result is processed in your mix. It's asking for trouble :-(

More to the point - if A/Ds are to be rated by how they respond when drastically over driven, that leaves the door wide open for A/D manus to 'exploit' you by deliberately building in errors and distortion into their converters, which you cannot remove when you don't want it.. This is a massive step backwards technically and undoes one of the valuable advantages of digital audio which is supposed to put you in charge :-(

[Waltz Mastering]

Originally Posted by Paul Frindle

I can't believe you guys are actually clipping A/D converters for sound effect!? And even going a far as rating them on the basis of how they respond to overload errors?

This is a very bad idea IMVHO - and what you will get will be a lottery involving what the converter actually spits out and what will happen when the erroneous result is processed in your mix. It's asking for trouble :-(

More to the point - if A/Ds are to be rated by how they respond when drastically over driven, that leaves the door wide open for A/D manus to 'exploit' you by deliberately building in errors and distortion into their converters, which you cannot remove when you don't want it.. This is a massive step backwards technically and undoes one of the valuable advantages of digital audio which is supposed to put you in charge :-(

I agree with you in principle P. but.... it is nothing new. It is has been a common practice amongst some ME's and some mixers for a long time.

It was not uncommon for engineers to push the limits in the analog domain, driving the tape into saturation... sometimes to a pleasing effect, so it would make sense that the limits and boundaries would be pushed in the digital domain as well. (to see what one can get away with)

For me, nothing is done by default, and once in a while driving the ad does work well. It is also one reason, in addition to a very good da, I have a very crappy da set up for monitoring... to hear what the consumer will hear.

I could be wrong but I had heard the DSM was a clipper as well...maybe not.. but anyway I've read on some forums where mixers are using clippers and limiters extensively in their mixes on drum busses, snare, and guitar channels etc... to in turn get their master louder....you know how it is on the internet, BS can spread like wild fire.