_______
I think the best thing to do is to divide
the sampling rate of CDDA by the
lowest and highest frequencies it was
intended to cover:


44,100 samples / 20Hz = 2,205 samples
per cycle.

44,100/1kHz = 44.1 samples per cycle.

44,100/20kHz = *ONLY* 2.205 samples
per wave cycle at that frequency. So
even I could see where things might
get a little dicey, sonically, up above
15kHz or so. That, combined with 16bit
depth and potential for lower amplitude
events in that part of the spectrum.