Dither me this Batman

Dither and bit depth and sample rate discussions should be banned from this forum LOL.


There is a great white paper on digital audio written by Dan Lavry around somewhere and also Nikka Aldrich wrote a book about it.

There is also a post somewhere where my ass got handed to me (nicely) by George Massenburg and Nikka years ago on bit depth.

I think, in all honesty, do a bit of searching around rather than about the millonth time this has been asked. It nearly always ends in a giant shit fight.

If you really want to go down the path of understanding this, get back to me and I will go find the links to the material that I used to learn about this stuff and post it to you
but be warned, it can get heavy going.

Or at least, rather than "i read something somewhere" post a direct link to the source material, and people can address it directly if they want in context of the accuracy of the statement you read.

Problem is, there is vauge understandings of this stuff, at best by the majority of people. Then also, now things are handled within plug ins and DAWS automatically, in some cases, that werent when articles were written.


I will say this. Categorically.

In every case I have heard presented, the difference between dither and no dither, it has been the least of the sonic issues in the example. Any time I have actually heard the effects of dither, I have had to crank the volume so loud
on my monitors as for it to be a moot point (thats moot not moops 8) )for all intents and purposes.

cheers

Wiz

Hi Henge

glad you took it like that, I had posters remorse 8) after hitting enter...

In that I hope it came across well, and not like a dick head.

cheers

Wiz

Don't know who would suggest something like that.
Leave alone dither until you export to the final bit depth. Leave everything else to the DAW.
Normally, the DAW does the following (if it is nearly intelligently programmed...):
Whenever a real quantization to lower bitdepth is done: Triangular Dither (TPDF), the only one recommended, if you will dither more than one time. This is maybe the case at freezing tracks etc.
Then, when you export to final media bit depth: Use a noiseshaped dither, if you have one at hand...
I use PowR-3.
That's all.
Most people seem to prefer the "Ozone"-dither iZotope MBIT+. Some freaks use even recorded analogue dither from Crane Song, they really sell an "analog dither CD". LOL. I have seen the noiseshaping curve somewhere, so i could do me a favour and "noiseshape" some white noise with the EQ to that curve and save me some bucks. LOL.
I guess this is some of the audio babble you can discuss to death over at the "forum we should not mention" slut..s and everybody would tell you you are an idiot if you are not using MBit+.
I would bet most of those discussing it could not tell an audible difference between freeware noiseshaping, preamp noise and the one they actually used on their own songs.
Dither and it's discussion is seriously overrated.
If you are a funny guy, just leave some preamp noisefloor in your tracks. May be sufficient as dither. lol.
Make an ABX test, if you want to make sure. I guess, this would clear it up for you. Then you probably never think about that again.

Best regards,
Martin

PS: I have been told "war" refuses to participate here.
You have been warned. If you now write something completely different, i would probably never answer. Hehe.

In modern DAWs and plugins, of decent quality, dither is applied automatically wherever it is necessary or "beneficial".

So, basically, don't worry about it.

If you're doing offline SRC or somesuch to digital audio files, and you are given the option by whatever software / process you are using, sure, use it.

But again, basically, don't worry about it.

Well, what does this tell us about, ehem, "Pro"-T.....?
No comment. Some software does have these kind of settings implemented by default or even configurable in the system settings (with the most appropriate pre-configuration done).
Well. Some. Maybe most. But not all. Obviously. Whatever does the trick...