Is dithering nonsense?

[brandy]

I hardly feel any differences between different types of dither when workin on a mastering from "that bandroom recording of that punkrockband from next door".

I would say that i even do not hear/feel ANY difference when using NO dither at all in such a scenario ;-)

But even here i will not miss to dither when going down to 16bit.

As it was said bevore:

If you deal with cleaner styles of music... maybe classical stuff or well recorded pop/jazz... You WILL hear artefacts in the reverb tails as well as in the fade outs etc.

Dither/noiseshaping "adds" something - so that you might listen to a "overemphasized" top end when using certain types of dither/noiseshaping and stuff.

But i would say that nobody cares about 0.001% more top end in a metal-scenario with distorted guitars and 10db+ EQ boosts all over the place ;-))


brandy

 

[ReliXKeepeR]

MetalWorks can you please post WAVs? Thank you!

 

[brandy]

I used to record in 16 bit but I got some songs to mix which were recorded in 24 bit and now I am trying to figure out which dither sounds best and if it makes a real noticeable difference....

Be carefull!

Even when you record in 16 bit (i assume that you use Cubase or Nuendo or similar) - the INTERNAL PROCESSING of that DAW software will be 32bit float. (talking abot Nuendo here).

So - if the project is set to 16 bit - you will record 16 bit audiofiles. Then the audio is processed in the mixer with 32bit float quality. So yu HAVE to use dither when you mixdown to 16 bit again in the end.

As i said above: Often you just hear no difference at all... But it is wrong to use no dither. If you have a L2/3 in the last slot you can use the built in dither/noise shape (wich is usually activated by default) - or you just put the UV22HR dither in the LAST plugin slot (post fader).


brandy

 

[Darkening]

I want to hear it also. They all sound the same to me.

Darkening - How loud did you have those files? Does that even matter? And what exactly are the differences in the two? Is it noticeable in a specific frequency range more or is it spread even over the entire range?

I listened to them on a pair of Blue Sky Media Desk 2.1's at low to moderate volume, definitely not loud. The room isn't even treated, but I'm sitting in the sweet spot. I certainly didn't have to struggle to hear it.

The undithered file sounds pretty harsh, which is what happens when you truncate 8 bits off the file without dithering. It isn't a specific frequecny range, it's the entire spectrum.

Record a couple tracks of acoustic guitars + vocals at 24-bit and flip that to 16-bit without dithering. Guitar distortion sort of hides the harshness, so listening to something clean might help.

I've used Digidesign's POWr dither type 2 and 3 on heavy rock songs and usually type 1 on softer songs and they sound fine to me, except for iZotope Ozone's Mbit + Ultra dither which I think sounds the best on all sources. Hearing the differences between different dithers is MUCH harder than hearing the difference between dithered and undithered, but not impossible.

Perform critical listening tests first thing in the morning, with fresh ears, or at least before you start tracking loud and mixing loud, etc. and start fatiguing your ears. Unfortunately, ears only perform at their optimum for a short while before they start to do their "auto-compression" thing.

 

[Radd]

Darkening - Wow! Thanks for all that info. So the dithered file sounds smoothed out compared to the undithered file? An analogy that comes to mind is ironing clothing.

 

[metalkingdom]

guy, seriously... i cannot believe this thread....

if your record and mix and 24 bit and don't dither before burning to CD the word length will be brutally chopped off at 16 bits and there will be audible artifacts.... c'mon now...

you may or may not notice... but that's down to how noisy your mix is to start with and/or how finely tuned your ears are.

Lol. I was scrolling to this post because I knew it had to be here.

 

[sorceron]

wow