🚀 go-pugleaf

Peter:
Thanks for all the nice examples of how to digitally create spectral folding.
As you show, there are all sorts of ways to create aliasing via
nonlinear DSP (I believe that I mentioned this as a possibility
previously in this thread).  I see that a low resolution A/D could be
modeled as a high resolution A/D followed by nonlinear DSP (e.g., sign(x))
to simulate the low res A/D and could, therefore, exhibit aliasing.
Unless we're dealing with a multirate system and/or one without a fixed
reconstruction LPF at pi, the maximum freq output is pi and attempting to
exceed pi will result in spectral folding.

I think we pretty much agree on what you wrote.  My original point was that
aliasing does not occur in the D/A process, but during sampling, resampling,
or nonlinear signal processing and, therefore, the D/A's output filter should
be called a "reconstruction filter" instead of an antialiasing filter.
Joe
-- 
.............................................................................
; Joe Campbell  N3JBC  jpcampb@afterlife.ncsc.mil  74040.305@compuserve.com ;
; My opinions are mine!   Happiness = Reality - Expectations, Click & Clack ;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~