RD>>  8-bit-sample = 16-bit-sample/256
 RD>> That ok ?
 MB>> i'd rather do:
 MB>> x=(x+128)/256
 EW> That's OK for signed values, not for unsigned's.
eeegegeeeeehhhh???
explain please. afaik, my method works for both signed & unsigned.
matju
--- Terminate 4.00/Pro
---------------