limiter/dithering

[Paresh]

Hi again - 2 questions: I got the classic master limiter for mix down. It says it will increase yr overall level (I always thought compression did that?). And it says if you mix dn to CD, apply a dithering plug right after. Does anyone know why & if this is really necessary? I see a dither plug in listed here but I know nothing about which one is best or how to use it? Thanks.

[multree]

well for normal demo use you can forget about dithering.....

Dithering is used when you convert higher resolution audio (24bit, 32bit) to 16bit CD qualit......

it inserts a little ground noise to compensate for the cut off bits....

at least that's what I understood....

[Yossarian]

I got the classic master limiter for mix down. It says it will increase yr overall level (I always thought compression did that?).

A limiter is a compressor, it’s just used with a bit more extreme ratio settings. Here’s a more in-depth explanation:

http://artistpro.com/index.php?module=P ... page_id=52

/Yoss

[Peel]

Limiting is a form of compression (specifically, compression with a high ratio).
Digital audio suffers from distortion at low levels -- ie, when only 1 or a few bits are being used to represent the signal. Dithering adds a small amount of noise which has the counter-intuitive effect of reducing this distortion, and increasing the apparent dynamic range.
Dithering is normally used when converting down to a lower bit depth (EG: 32 or 24-bits down to 16). The basic rule of thumb is, don't covert down to a lower bit depth until you must, and when you do, dither. Dithering should be the very last signal processing you do (ie should be last in your signal chain, after your master fader, if your master fader isn't at 0dB). Dithering is only beneficial when used this way.

[M'Snah]

http://artistpro.com/index.php?module=P ... page_id=52

Thanks for the URL. Although short pieces of text, it's very useful information for me.

I read that it's common to use compression when recording, to get a stable / hot signal? I thought it was best to record the signal as pure as possible for further processing... I learn something new everyday

--HansM

[Yossarian]

I read that it's common to use compression when recording, to get a stable / hot signal? I thought it was best to record the signal as pure as possible for further processing... I learn something new everyday

I think this is becoming less critical as people start using higher bit rates. When recording at 24 bits or better you have so much headroom that you don’t have to worry too much about keeping the level as hot as possible. Recording with compression in this case becomes an option, not a necessity. At least this is my understanding – I’m not an expert by any means.

/Yoss

[pHz]

with regard to dithering ... check out the izotope / ozone guide ...

http://www.izotope.com/products/audio/ozone/guides.html

slainte rob

[Kriminal]

Hi again - 2 questions: I got the classic master limiter for mix down. It says it will increase yr overall level (I always thought compression did that?). And it says if you mix dn to CD, apply a dithering plug right after. Does anyone know why & if this is really necessary? I see a dither plug in listed here but I know nothing about which one is best or how to use it? Thanks.

dont waste time dithering, keep your audio at 16/44 and you will be absolutely fine. The rest is for bats and dogs and anal boffins

[kurasu]

The rest is for bats and dogs and anal boffins

or for people who actually gives a shit..

[chagzuki]

Dithering only makes a difference on extremely low level 16bit signals, something like -70db, I can't remember exactly. It's really beyond the point of making any difference in most conceivable circumstances.

[XDVarenkor]

Recording with compression in this case becomes an option, not a necessity. At least this is my understanding – I’m not an expert by any means.

There are other reasons to use compression during recording - for example, the sound of a hardware valve compressor is hard to emulate otherwise (sit down at the back there, VintageWarmer! )

[Kriminal]

The rest is for bats and dogs and anal boffins

or for people who actually gives a shit..

No, its for ppl who like to think they can hear the diff/quality. I can put up 3 files of diff resolutions if you want, and i bet you couldnt tell me which is which without looking

I can even render you a 32bit IEEE/384kHz EHP and put that to the test if you are game

[IIRs]

dont waste time dithering, keep your audio at 16/44 and you will be absolutely fine.

There's nothing wrong with being ignorant, but don't offer advice if you know jack-shit.

It is not possible to "keep your audio at 16/44" because every process you do (including mixing two 16-bit files together btw) will increase the wordlength of the audio signal. This is why your host almost certainly uses 32-bit floats for the mixing engine regardless of the bit-depth (bit-rate refers to audio compression such as mp3) of the original recordings.

Your final mixed signal will therefore be 32-bit anyway and will need dithering down to 16-bit if you wish to keep as much information as possible.

Dither should be used whenever the wordlength is reduced, even if going from 32 or 48 bit to 24-bit. Dither basically makes the lowest bit of the reduced wordlength signal toggle on and off in a way that statistically re-creates some of the information that would have been lost otherwise. In other words it can actually increase the apparent bit-depth of CD audio, some say to 18-bit or better.

If you can't hear a difference between truncated 16-bit audio and dithered 16-bit audio, it either means your recording had no subtle low-level detail to start with, or your monitors are not accurate enough..

..or you have no ears.

[Peel]

It's very easy to hear the difference between dithered & non-dithered audio at 16 bits during fadeouts, reverb trails on final notes, etc. Hardly "beyond any conceivable circumstance."
And how big the difference will be between 16-bit & 24-bit really depends on what you & your software are doing with the audio. For example, there will be a very significant difference between 16-bit & 24/32-bit audio if you're doing destructive processing.

Compressing during recording can be a good idea, particularly if you're recording at 16-bits. But this assumes you're using a hardware compressor before the analog-digital conversion of your soundcard. Using a software compressor for this purpose is pointless (from a technical sound quality point of view -- it might still sound nice).

[Yossarian]

There are other reasons to use compression during recording - for example, the sound of a hardware valve compressor is hard to emulate otherwise (sit down at the back there, VintageWarmer! )

Exactly - which is why I said it's an option.

/Yoss