In article , Chris Hornbeck
wrote:

And to make the whole thing even soupier, dither can theoretically
be removed after DAC

If you're talking about subtractive dither, that's pretty impractical
after the DAC because you need to know exactly what the dither signal
was and reproduce it precisely, but inverted. Dither needs to be random
(or pseudo-random) by it's very nature which makes the technique quite
challenging under non-labratory conditions. How will the CD player at
Joe Schmoe's house know what dither you used? If you were to include a
key or "antidote" with the recording, the encoding of the dither
"antidote" eat quite a bit of storage space (it would be continuous
alongside the audio), and of course you'd need new hardware that
supported the subtractive technique and could make use of that
"antidote". We won't even get into the potential production problems
when you start editing and processing. The technique is more practical
if used in an ADC where the designer has control over it as a closed
system, but even there I'm not aware of a practical example in
widespread use in audio production.

If that's not what you were referring to, then just think of this as an
interesting (or not so interesting) aside...

--
Jay Frigoletto
Mastersuite
Los Angeles
promastering.com