[David Nebenzahl]

On 1/26/2010 9:14 PM Mr.T spake thus:

"David Nebenzahl" wrote in message
s.com...

No, by empirical evidence I meant some numeric value of comparison. Say,
maybe, number of samples changed between original and re-encoded stream,
or some such, as a rough measure of corruption.

You could have 100% of samples changed without audible corruption, (or 1
sample changed can be audible) so I can't see what you expect that to tell
you?

Well, that's true; just counting changed samples won't tell you
anything. But I'm sure there must be some way to quantify the level of
corruption, or more properly the degree of difference between a
re-encoded file and the original, based on what would actually be
audible rather than just raw numeric differences between samples. It
wouldn't be trivial to calculate, that's for sure.

Anyhow, at this point this is just an academic question and we've moved
way beyond the OP's issue.


--
You were wrong, and I'm man enough to admit it.

- a Usenet "apology"