The Logic Off Topic list

> From: Spectro <spectro@...>
> 
> What is the justification given for such a high rate? and what
> exactly would one be sampling at that rate? I'm curious...

Check out this months' AES journal for the whole scoop... basically the guy
plotted aliasing noise vs. sample rate vs. clip level for a dual-tone input,
and the aliasing dropped to an acceptable level (around -120dB) at about
6MHz sample rate.

Whenever you clip a waveform (with a non-linear mapping function - there are
a couple of plugins that you can do this with), you introduce distortion,
sidebands, however you analyze it.  Simply put, with a simple clipper, you
get odd-order harmonics with amplitudes falling proportionally.  Those will
then alias based on their frequency relative to the sample frequency.  So if
you raise the sample frequency, by the time you get to harmonics that start
aliasing, their amplitude is sufficiently low to be inaudible.