[sdiy] SSM2040 filter question

[Antti Huovilainen]

On Wed, 12 Nov 2003, Magnus Danielson wrote:

> You can pre-distort the responce to get a better result, but better than
> that is to take the original design limitations and apply them on a
> modern filter synthesis rather than taking an old one and map it over.

What do you mean by original design limitations?

> Also, just think about the CV-control aspect of things and you start to
> shiver.

That's a problem if you apply bilinear (or other) z-transform to the
complete filter transfer function. You also lose nonlinearities. So one
must separate the filter sections and handle each one individually.
The tuning is then a simple function that can be easily handled with
interpolated table lookup.

> > IMHO the sample rate must be raised so that the phase properties
> > inside the bandwidth of interest are not affected.
> Exactly! I have long argued for this on the same grounds!

Do they actually matter?
Phase matters for resonance but other than that I don't think it matters
at all.

Even 2x oversampling corrects much of the amplitude error and I don't
think ear is That sensitive to it, especially as it occurs at such high
frequency. In practise I think the lower limit to oversampling is due to
aliasing caused by the filter nonlinearities.

Antti

Give a man a fire, and he'll be warm that day,
Set him alight and he'll be warm for the rest of his life