scream

[mbartkow at ET.PUT.Poznan.PL]

> So far, this seems like a situation where the digital version of a
> circuit is unable to duplicate an analog circuit. 

It is almost obvious, since nonlinearities produce high-frequency harmonics
which cause aliasing in sampled systems.

> I suppose I could
> normalize the input to the filter, but this would be difficult for live
> signals.  Anyone know of any good bandwidth limited, amplitude limited
> nonlinearities that would work in a feedback loop for digital systems?
> Sounds easy enough. ;)

Chebyshev polynomials do the trick. They are both band limited and amlitude
limited (at least in the -1..1 range of arguments). Moreover, they are very
nicely "twisted", especially the higher order ones. That's why I mentioned 
them earlier and I am so desperately looking for a circuit of Chebyshev
wave shaper.

> P.P.S. What do the diodes in the Q feedback loops of some state-variable
> filters do? How does it change the sound?

Isn't their purpose to introduce clipping in order to prevent excesive 
build of self oscilation ? I think they are mostly responsible for the
filter "bleeding" when overdriven...

regards,

--

Maciej Bartkowiak
========================================================================
Institute of Electronics and Telecommunication     fax: (+48 61) 8782572
Poznan University of Technology          phone: (+48 61) 8791016 int.171
Piotrowo 3A                             email: mbartkow at et.put.poznan.pl
60-965 Poznan POLAND               http://www.et.put.poznan.pl/~mbartkow
========================================================================