dithering algorithms and resampling

POW-R 1

POW-R 2

POW-R 3

Waves L1

Waves L2

iZotope MBIT+ (ExtraBit/MegaBitMax)

i'm sure there are tons more...

i've been using mbit+ just because it's there... and i like the izotope ozone plugin for compressing and such

but i'm not sure i can tell the difference between mbit and pow-r

i haven't tried those offered from the Waves plug-ins though

are there any that you guys don't like and do you know any disadvantages to mbit or pow-r?

----

also, resampling from X -> 44100

and are there better / worse ways to do that?

i usually just do the slowest setting in samplitude

and.. before or after dithering?

While we're on the topic of dithering... how bout an explanation about what it's for? I think I have an idea of what it is and why you'd use it, but lets have an in depth look at just so I can know that if what I know I should be knowing. you know?

i'm no expert but here's a collection of my assumptions

when looking at a waveform

sample rate is the horizontal distance between samples

bit depth is the vertical distance between samples

maybe compared to film...

sample rate :: framerate

bit depth :: frame resolution(?)

often people record at higher bit depth and sample rate than their intended target, mix / effects in that resolution and then once the final mix is done they'll adjust it to fit their target medium

often CD

sample rate 44.1 khz

depth 16 bits

but going from 96khz / 24bit to 44.1khz / 16bit you're going to lose data and this can cause artifacts and oddness with your ears.

if you're familiar with image editing and GIFS think about a GIF with a 256 color palette and then reduce it to.. 32 colors or something. if you just straight up chop off those other colors it can get a little blocky and distorted.

but if you use the different dither types like noise/pattern/diffuse you can make the degradation appear more graceful.

so.. same goes with audio and these different methods of dithering. they each have ways of dealing with the noise, artifacts, and what-have-you that arise and try to adjust them in such away as to be inperceivable to human ears.

Alright thanks, I had it right :D

I've never experimented with dither much, as I usually mix down to tape for various reasons. But thanks to you, I'm gonna do some dither research! 'Twould be good knowledge for making samples anyway...

though, I would explain sample rate and bit depth a little differently. Sampling rate being the resolution at which the frequencies are recorded, and bit depth being the resolution of amplitudes recorded. I guess your explanation is easier to understand though.

dithering is low level noise added the signal to modulate the least signifigant bits before they are truncated

chloroplast has it

sample rate is very much like the framerate

sample rate affects your frequency spectrum wheras bit depth affects your amplitude(not overall but accessable points between)

my favorite dither is the apogee uv22hr

technically noise shaping with autoblack

basically is dither with an algorythm placed over it to move the noise around the pass band and the ability to shut of when unnecessary

yeah i ain't the pickiest mofo in the universe, but since I've used wavelab I've quite enjoyed UV22.

still i dont' give a fuck and use any ditheringi can get usually

when i took a mastering class we used SADiE (a mastering DAW)...UV22 seems to be industry standard for dither...

hey guys, would you take this blind test maybe?

http://www.24-96.net/dither/blind_shootout.htm

  Quote

All files have been dithered to 16 bits from the same original 24 bit file. 54 dB of gain has been applied after dithering.

21 different sound files each using different dithering methods and different settings.

i'm curious which you'll choose

i narrowed down to:

megabitMax Ultra

Waves Ultra noise shaping

megabitMax High

and chose mega bit ultra

it tells you, after you choose and finish the test, which are which

(it helps to save them all on your drive first and playlist them so you can jump back and forth between the apparent 'ties')

also curious if you see any problems with the test because the results seem exceedingly in favor of one of them.