[Scott Dorsey]

wrote:
Scott Dorsey wrote:

The thing is, the nonlinear encoding used by ADPCM isn't
really either of those things.

Then what is it?? Come on, I'm black & white here, no time
for fuzziness, lol!

It's a nonlinear encoding. Amplitude step is log(n) instead of n, so
the size of step 0 is smaller than the size of step 127. If you look
for an explanation of u-Law online, that is the most common method used
in the real world. But it's still straight PCM.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."