Dithering - 24bit to 16bit - POW-r #1, POW-r #2 or POW-r #3 - or no dither?

First off, if I were you I'd look into what Dithering actually is. There are many resources on what this stuff does.
Basically it's the process of adding a small amount of noise, or hiss to the sound so that the truncation from 24bit to 16bit doesn't lose as much quality as it otherwise would.
It's something that only affects very very quiet sections of music and slow fade outs, which you don't really seem to find anymore anyway.

The different types of Pow-r dither are simply noise shaping so it's trying to make the noise added less apparent to the human ear, while at the same time achieving the goal of smoothing out the truncation so you don't hear it. There are different types to fit better behind different subject matter. The recommended uses are detailed in the logic manual, so have a look there.

Also if there noisefloor of your recording never falls below the dynamic range of 16bit then you strictly speaking don't need to do it anyway.

Also if there noisefloor of your recording never
falls below the dynamic range of 16bit then you
strictly speaking don't need to do it anyway.

And in that case, you could also record in 16bit without quality loss.

POW-r #3 is very highly regarded among many top Mastering Engineers.

Advice to you is do bounces of the same music and listen to the differences in the various types of dithering and use the one most appropiate for the music.

You should always dither when reducing bit depth.

Here is some detailed information regarding dithering:
Dithering Article by Bob Katz



Dual 2GHz G5, Apogee Rosetta 800, RME 9652, Focusrite Red Series Mic Pre, HR824's Mac OS X (10.4.8)

And in that case, you could also record in 16bit without quality loss.

No quality loss due to dithering, but other aspects of your recording will be compromised.

to spirit303:

Regarding listening to the differences of dither. Chances are, you won't hear any difference. The differences are indeed very subtle.

My advice would be to pay close attention to your "gut reaction", when listening to the various dithering algorithms. It's my belief we can sometimes tune into these subtle differences, though we may not be able to quantify why we think it's better.

POW-r is for eliminating distortion and artifacts when going from 32, 24, or 20bit (anything) down to 16bit.

I'm probably going to be railed for this, but in a nut shell IMO:

#1- simple for spoken voice
#2- more complex good for rock band
#3- most complex good for symphony orchestra or other complex situations



Dual 2Ghz Powermac G5 -- 15" Ti Powerbook G4 867Mhz Mac OS X (10.4.10) Logic 7.2.3 Ubuntu PPC

hi, agreed.
i don't use any, if i can get away with it.
cheers, david r.