Re: 16 bit vs 24 bit Information Theory Hypothesis

["Lou Judson" inaudio]

Something unclear on the concept. It is not "upsampling"or
"downsampling" - those terms refer to sample rate, not bit depth.

You "dither it down" (or truncate it) to 16 from 24, and if you
import a 16bit file to a 24 bit app, it simply adds 8 zeros. But when
you do even the simplest proecss on it - even change level a tenth of
a dB, it will be 24bit and need dither to go to 16 again.

It is not conversion at all. Wrong term!

<L>
On May 26, 2007, at 6:33 PM, Tim Nielsen wrote:

> Except! When you convert you sample from 16 bit to 24 bit for
> processing, the upconversion must 'make up' the missing information.