[linux-audio-dev] Re: Cheby amp code

[Tim Goetze]

Steve Harris wrote:

>On Sun, Nov 03, 2002 at 07:48:55 +0100, Tim Goetze wrote:
>> >OK, why do they use two shapers? Or is it one cheby and one non polynomial?
>> 
>> two chebyshevs, blend factor depending on incoming amplitude.
>
>Thats interesting, cos liiing at your waterfall plot it doesn;t seem liek
>that should be enough.

afai understand the paper, they've been doing this as an academic 
exercise. ivory towers and so on ...

>I odnt think that Line 6 and co. are using chebys, I think there using
>something more like our first attempt. That's not based on any knowledge
>or anything, just a hunch.

yep, it seems unlikely the more i think about it. the peak
problem when lots of strong harmonics are generated could be 
tackled with a phase-shifting filter. but we'd still need to split
the input into multiple bands to come up with purer sines (i've
done some more experiments with two mixed sines and a chebyshev
shaper: the output is badly garbled, with unpredictable peaks).

but pod distortion waveforms don't look like they use chebyshev
polynomials at all.

tim