28 Oct
2002
28 Oct
'02
12:58 p.m.
Steve Harris wrote:
On Mon, Oct 28, 2002 at 01:30:49 +0100, Tim Goetze
wrote:
yep, that's what i expect, too.
it's an interesting pov though, in fact
i've been thinking
about algorithms that work like IIR filters (that are often
used for HP/BP/LP) with the addition of nonlinear terms for
recreating the tube distortion. the maths of such a thing are
not easy to grasp though.
In my (limited) experince its really hard to make them stable. But you can make interesting non-linear filters by using different
parameters for the
different halves of the wave cycle. This is what makes the gong sound
gongy (waveguide_nl.h). It could also apply to cone modelling, as the
characteristics of the deflection in the positive and negative directions
will be a bit different.
it's interesting to know that valve preamp stages should have
a small bias (to + iirc), and you can in fact observe this in
the amp output i've recorded.
To make a nonlienar cone filter you could either use
different IIR
parameters for signal sign, or you could us different parameters for
different amplitudes (or both).
i'm still trying to model the hard-driven valve in the time
domain. the more i look at it, the more i think the 'building of
potentials' i've tried to describe comes near it. though it would
be nice to base it on hard facts about valve amp behaviour, too.
interestingly, the sine, shaped by the hard-driven amp, looks
a lot like the plots of 'ion conductivity' and 'Aktionspotential'
from human heart analysis.
unfortunately they don't give a formula in the book i have on my
lap. any heart surgeons around? ;)
Different amplitudes would be equivalent to how you
model nonliear systems
with FIR impulses - I can't believe I didn't think of it before.
in fact you described something along these lines a few days
ago. ;)
tim