For a bandpass filter, wouldn't a lowpass and highpass in series be
ideal On the other hand I've been thinking about a variation on the
classic diode ladder design:
4 6db lowpass normalled in series and with a main frequency control
for all four as well as individual frequency potentiometers. There
would only be two freq cv inputs: all and spread

As well as individual outputs one could provide a mix out with
polarizers for every stage.

This configuration would allow for every possible filter type and
slope including tb-303 emulation, by tuning stage 4 one octave above
the rest. What would be really cool is if it would have morphing
capabilities like the a-107 (I think, don't remember the numbers by
heart). But this filter would be superior (custom types, slopes, no
switching clicks etc)

Cheers

26 apr 2010 kl. 10.09 skrev < yahoo@... >:

> > Creating a filter slope variable is quite possible. These are the
> > commercial modules I know of.
>
> The main question is: do they pan/morph between the 6/12/18/24... dB
> outputs
> of a usual filter design or do they use a filter circuit that has
> indeed the
> filter slope as an additional parameter (I mean if it can generate
> really
> e.g. a 9dB slope, not just mixing a bit the 6dB and 12dB filter
> stages). I
> cannot imagine such a filter based on an analog design, but who
> knows ...
>
> In any case 'd be very interested if someone knows details about
> such an
> analog filter design.
>
> For bandpasses the situation is a bit different as the slope of a
> bandpass
> is also affected by the resonance: the bandpass peak becomes
> narrower as the
> resonance increases. If the filter design has a constant level at
> the filter
> peak the slope of the bandpass becomes higher as the resonance
> increases.
>
> Best wishes
> Dieter Doepfer
>
>

[Non-text portions of this message have been removed]