[sdiy] jitter in oscillators for music purposes

[Sean Costello]

When working on models in the past (both analog and physical), I have
developed the high cost, high quality models first. Once you have a good
model as a basis, you can start optimizing the code, simplifying the model,
etc. In general, the initial model has been based on an accurate physical
realization, and the optimizations are attempts to get similar PERCEPTUAL
effects at lower cost.

Sean Costello

----- Original Message ----- 
From: "Czech Martin" <Martin.Czech at Micronas.com>
To: "Sean Costello" <seancostello2003 at comcast.net>; "Antti Huovilainen"
<ajhuovil at cc.hut.fi>
Cc: "Colin f" <colin at colinfraser.com>; <synth-diy at dropmix.xs4all.nl>
Sent: Monday, July 19, 2004 1:42 AM
Subject: RE: [sdiy] jitter in oscillators for music purposes


> Yes, interesting. In order to not mix up effects of the
> time discrete computation with the phase and comparator
> noise modulation some computational overkill can be
> accepted.
>
> Later on, it's time to save.
>
> m.c.
>
> > As far as a simple approximation, how about adding noise
> > directly to the
> > phase increment value? This will give you phase modulation,
> > as well as a
> > small amount of amplitude modulation (due to the fact that
> > the current phase
> > + phase increment + noise may wrap around, whereas current
> > phase + phase
> > increment would not).
> >
> > I am interested to hear about the perceptual effects of the precise
> > simulation of jitter in the oscillator. However, once the
> > perceptual effects
> > have been worked out, it seems like it would make sense to
> > try approximating
> > the effects with a simpler noise function, in order to be
> > able to use the
> > results in a real-time analog synth emulator.
> >
>