[Mark]

So that means the beats and whistles I have seen, and
heard, *when I feed tones in the range of 10 kHz to 20
kHz *into a low cost unit are not due to "simple
aliasing" *of the input tone but instead are due to some
other higher order phenomenon, probably related to
aliasing of the harmonics..

Please identify the product in use and the operating environment in greater
detail.

Phillips CDR560 CD recorder fed with an sine audio signal
generator...

I will try to get to the bench this weekend and repeat the experiment
and provide more details..

See also:
http://src.infinitewave.ca/

This is a comparison of Sample Rate converters, but the issues of the
aliasing and unwanted beats in sample rate down conversion are similar
to what can happen in an A/D.

Set the test result button to "sweep" and comapre the Adobe Audition 2
with and without pre/post filter for example to see a similar (but not
the same) effect.

Anytime1.0 and Sadie look like an examples of bad implamentation.
SSRC looks like an example of good implementation.
And it is harder to get it right at 44.1. compared to 96.


Re 96 vs 44.1.... my conclusion from looking at this data is: 44.1
__CAN__ be just as good as 96 IF it is implemented correctly.....but
if it is not implemented correctly, it will be a lot worse...

But if you have a good piece of gear that with a good implementation
at 44.1 , then no problem...

So as usual the bottom line for me is trust but verify, if you have
tested your gear at 44.1 and it works well, then good... but unless
it has been tested, then anything goes. The same is probably true for
96 but perhaps to a lesser degree.

Mark