Chaotic Oscillators (over-blown flutes etc)

[Martin Czech]

>  I'm just wondering if there
> might be a way to simulate the tube and the reed somehow.  The tube is
> like some kind of sympathetic resonator (VCF with rez), the reed some
> kind of relaxation oscillator (saw VCO?), but how to simulate this is
> a mystery to me since obviously most of us have played with a saw VCO
> and a VCF with rez with various parameter settings and it doesn't
> usually sound very much like a clarinet or any other woodwind.  It's
> like the vibrating column of air is a filter that confines the reed's
> vibration to a narrow range of pitches, almost like a PLL.  I'm going
> to keep thinking about this, but I doubt that any circuit I design
> will sound like a clarinet.  (hopefully <g>)

I think  a part of the magic is the feedback between the tube conus and
the reed. The pressure wave from the reed (reed is now open)  moves
forward down the tube and is reflected at the hole at the end. Running
back it influences the reed (shut it off). This way there is a strong
nonlinear feedback between reed and pressure wave in the tube.
Therefore it should be a good idea to model the tube with a delay line
(with a kind of lowpass filter for high frequency damping) and the
reflected wave should sync the oscillator or cause some other kind of
phase jump.  Since the tube is able to carry  a lot of wave modes it is
possible to overblow the instrument.

Thus the tube cannot be modeled with a simple lowpass filter.

m.c.