Deliberately introducing clipping?

  • Thread starter Thread starter brassplyer
  • Start date Start date
brassplyer

brassplyer

Well-known member
I've always been under the impression that avoiding clipping is a basic tenet of digital audio. Looking at this mastering tutorial, if I understand what this guy is saying, he deliberately introduces a limited amount of clipping - 2:30 "I wanted to get a little bit of clipping on the A to D".

??

Yoad Nevo: Mastering with Waves - YouTube
 
Some people do that. There is a sound to it and some people, as an artistic decision, intentionally clip the audio to get that effect.

Different systems/pieces of equipment clip "nicer" than others. ymmv.

Also, if you are just clipping off the snare drum transient (for example), the distortion from the clipping will be pretty i audible.
 
brassplyer said:
"I wanted to get a little bit of clipping on the A to D".
Click to expand...

I'm pretty sure modern ADCs will clip the analog stage before they clip the digital stage. So it would be the sound of that analog clipping he likes. You can clip digitally ITB easily enough, but that sounds different. I've seen audio from major releases with obvious clipping of one kind or another.
 
bouldersoundguy said:
I'm pretty sure modern ADCs will clip the analog stage before they clip the digital stage. So it would be the sound of that analog clipping he likes. You can clip digitally ITB easily enough, but that sounds different. I've seen audio from major releases with obvious clipping of one kind or another.
Click to expand...

Indeed they will. The analogue drive stages to the converters HAVE to self limit before the protection diodes kick in in the A/D chip otherwise you are in danger of blowing them.
This is usually arranged by having the drive chips powered from the same rail as the A to D device.

Dave. (As ever. Mr Self tells us how and why!)
 
ecc83 said:
Indeed they will. The analogue drive stages to the converters HAVE to self limit before the protection diodes kick in in the A/D chip otherwise you are in danger of blowing them.
This is usually arranged by having the drive chips powered from the same rail as the A to D device.

Dave. (As ever. Mr Self tells us how and why!)
Click to expand...
Wow I've never read what actually goes on in the front end like this.
If I'm getting it right, diodes in/on the chip might be there to protect even more delicate circuits down stream? But then logically I suppose would have their own physical limits, thus the voltage rail staging from higher more gross overloads?

If you don't mind- a new question? I've been very curious of late about clipping- but at my master file's output and the D/A. Basically, I found cases where I actually liked clipping (the snare for the most part) better than limiting just below 0dBFS.
I never quite understood though in that scenario if what we're hearing is the file' ('math?) clipping, or the analog parts of the D/A?
 
mixsit said:
Wow I've never read what actually goes on in the front end like this.
If I'm getting it right, diodes in/on the chip might be there to protect even more delicate circuits down stream? But then logically I suppose would have their own physical limits, thus the voltage rail staging from higher more gross overloads?

If you don't mind- a new question? I've been very curious of late about clipping- but at my master file's output and the D/A. Basically, I found cases where I actually liked clipping (the snare for the most part) better than limiting just below 0dBFS.
I never quite understood though in that scenario if what we're hearing is the file' ('math?) clipping, or the analog parts of the D/A?
Click to expand...

Can't help you much further than the analogue A/D interface I am afraid. I know the chip has some protection diodes across its input but of course inside that SM package they are going to be mighty wee! The drive is fed in via about 240R- 1k resistors and so the voltage at the driven end must not be too big otherwise too much current will flow. Can't make the feed resistors any bigger or they won't charge the devices inside the A/D chip. So, the supply rail is limited.

The "math"? Getting a bit beyond me now! I know most audio software uses Floating Point Math. You did that at school? 1,400,000 is 1. 4 times ten to the six? So 32 bit software can handle almost limitless numbers and internal mixers cannot clip...BUT the "number" CAN get too big for the output stage which is limited in even the best pro hardware to around +24dBu or around 15volts rms.

But, if you want the REAL SP on this stuff check the back issues of Sound on Sound.

Dave.
 
ecc83 said:
.. The "math"? Getting a bit beyond me now! I know most audio software uses Floating Point Math. You did that at school? 1,400,000 is 1. 4 times ten to the six? So 32 bit software can handle almost limitless numbers and internal mixers cannot clip...BUT the "number" CAN get too big for the output stage which is limited in even the best pro hardware to around +24dBu or around 15volts rms.

But, if you want the REAL SP on this stuff check the back issues of Sound on Sound.

Dave.
Click to expand...
Thanks. Yeah, that's where I sort of loose it is the mix math is 'float point ie deep as far as dynamic range, but what is happening when it gets sent to the D/A.
 
Doubly fo sho!

Clipping as being discussed here is an artistic decision just like the use of reverb or compression. If it suits your style and gives you the sound you want, go for it.

However, as always the caution is (just like any other effect) if you add it before recording you're stuck with it and can't change your mind later. Be darn sure it's the sound you want before committing it to disk.
 
Yikes!

I just listened to samples on the site - his work is a little too loud for me
 
Once you know the rules I guess you can break them if they achieve your production or artistic requirements. However in context of mastering clipping is kind of "old hat". With mastered for iTunes and volume normalization likely to become more and more prominent in relation to music consumption there is little to gain and a lot to lose. I would avoid it personally. I rarely clip audio these days, unless a client is very insistent on a extremely hot master for whatever reason they have got into their heads. I speak to clients and understand their needs as much as possible and help come to a sensible optimized end result, whatever that may entail.

cheers

SafeandSound Mastering
Audio examples and mastering FAQ video
 
I agree on the artistic usage of clipping. The clipping characteristics sound different on a digital audio signal platform than an Analog! In my past project where certain frequencies of my production needed that "good" distortion, or even that full loudness, I'd record the track material or full mix to Tape! Pump up the levels and record! Ill then take my recording on tape and record that back to Wave! Much pleasing sound in the high levels! I see alot of modern producers now peaking their work at 8db rms! In the Digital world, and this is there final mix down, was never sure why they'd go that high! Say you take that recording, export it to mp3 and play the material in your car stereo or iPod or iPod dock! The one main thing you notice is there's no headroom because the signal is extremely hot, things like snares and cymbals and hi hats have noticeable clipping! I could understand maybe the bassline or desired distortion from a midrange frequency like guitars, or even the voice! But the whole mix isn't too pleasing!
 
You must log in or register to reply here.
Back
Top