[sdiy] Craig Anderton's Multiple Identity Filter

Fri Oct 28 03:16:54 CEST 2016

On 28 October 2016 at 06:42, David G Dixon <dixon at mail.ubc.ca> wrote:
> Yes, the Xpander is what I meant.

Actually, my mistake! As Brian pointed out the Matrix 12 is a double
Xpander, I just wasn't aware it existed.

> However, I'm doing it somewhat
> differently from Richter.  The way he did it only gave certain responses.
> We tap into the feedback loop as the "input" signal for the summing, which
> opens up a lot of proper highpass and notch modes otherwise unavailable.
>
> As far as the strange behaviour you speak of, I'm not sure I see that.  I've
> got an Excel model which allows me to plot the transfer functions of every
> mode based on the gain factors, and with a slider that changes the resonance
> from 0 to 99% of self-oscillation.  I don't actually see the behaviour you
> mention in the bandpass modes.  In highpass modes, there is about a 3dB dip
> at about log w = 0.25 (where log w = 0 at the cutoff frequency) at
> self-oscillation, but I'm not sure that this isn't actually normal highpass
> behaviour.

Try feeding the input signal with the resonance subtracted to the base
of the first stage cap, then each subsequent stage's output to the
base of the next cap, and plot those paths, you will see the "correct"
low pass shape mirrored. If you check the schematic of the Moog high
pass filter this is what is going on, although I'm not sure they do
this for their "Ladder" module as I've never seen a schematic for
that, or heard one.

> As far as the attenuation of frequencies below the cutoff with increasing
> resonance, even the lowpass outputs (the natural outputs of the filter) do
> this.  This is normal behaviour for the cascaded-stage filter.

With the "correct" responses high pass attenuates the high frequencies
with increasing resonance, the low pass the low frequencies, and the
band pass stays put, the attenuation remains constant either side of
the cutoff and increasing resonance only increases the resonant peak
amplitude.

Anyway, the main point of my bringing this up was to point out that to
get "correct" responses you actually need to change both the position
the feedback is sent and it's polarity, neither of which I can see in
Craig Anderton's schematic, so you'll get a mess if you don't to that,
the output summing is much easier to do and you don't have to change
the resonance path, but the responses are a bit odd when doing this.

Cheers,

Andy

>
>> -----Original Message-----
>> From: Andrew Simper [mailto:andy at cytomic.com]
>> Sent: Wednesday, October 26, 2016 6:58 PM
>> To: David G Dixon
>> Cc: KA4HJH; sdiy
>> Subject: Re: [sdiy] Craig Anderton's Multiple Identity Filter
>>
>> Hi David,
>>
>> As far as I remember the Oberheim Matrix only had a fixed 4
>> pole cascade type low pass filter, did you mean the Oberheim
>> Xpander? If so then yes this is the easier way to generate
>> different responses by taking weighted sums of the output of
>> each stage, but you don't get the "correct" responeses, you
>> get sags above the cutoff in the band pass and high pass
>> modes, and amplitude attenuation below cutoff with changing
>> resonance in all modes, which kind of sounds like changing
>> the cutoff upwards with increasing resonance in band pass and
>> high pass modes, but the resonant peak stays put.
>>
>> Cheers,
>>
>> Andrew
>>
>> On 27 October 2016 at 03:23, David G Dixon <dixon at mail.ubc.ca> wrote:
>> > I've got a filter design called the "Polaris" that gives nearly 30
>> > different useful filter modes in a much simpler say than
>> the MIF, and
>> > doesn't need a Curtis chip to do it.  Intellijel sells it
>> for $229.
>> > If you want to check it out, go here:
>> > https://intellijel.com/eurorack-modules/polaris/  It's based on a
>> > Roland-esque cascaded-stage four-pole filter, and takes inspiration
>> > from Grant Richter's old Oberheim Matrix multimode filter
>> design, but
>> > it goes a fair bit further than that.  The mode switching
>> is done with
>> > a microcontroller generating a 17-bit code which controls
>> five analog
>> > switches, which I still don't know how to do, so I don't
>> have a handmade version of it yet.
>> >
>> >> -----Original Message-----
>> >> From: Synth-diy [mailto:synth-diy-bounces at dropmix.xs4all.nl]
>> >> On Behalf Of KA4HJH
>> >> Sent: Wednesday, October 26, 2016 8:55 AM
>> >> To: sdiy
>> >> Subject: Re: [sdiy] Craig Anderton's Multiple Identity Filter
>> >>
>> >> > On Oct 26, 2016, at 4:37 AM, Andrew Simper
>> <andy at cytomic.com> wrote:
>> >> >
>> >> > From what I can see with Craig Anderton's filter it
>> won't actually
>> >> > give the "correct" responses for all the different filter
>> >> types unless
>> >> > there is a way to switch the polarity of the resonance, and
>> >> also where
>> >> > the resonance is added. Am I missing something? Sorry if
>> >> I've posted
>> >> > all this stuff before.
>> >>
>> >> This is one of the reasons I reposted the link. I remember some
>> >> discussion about this and wondered how Craig's design compared.
>> >>
>> >>
>> >> > On Oct 26, 2016, at 12:30 AM, Doug Terrebonne
>> >> <dougt55 at yahoo.com> wrote:
>> >> >
>> >> > The MIF article has been on synthdiy.com for a long time -
>> >> > http://www.synthdiy.com/files/2002/MIF.pdf
>> >>
>> >> I didn't even know it was there.
>> >>
>> >>
>> >> > On Oct 26, 2016, at 5:43 AM, Roman Sowa <modular at go2.pl> wrote:
>> >> >
>> >> > Why use predatory file hosting sites while there are sites
>> >> like synthdiy.com or synth.net, which AFAIK are managed by
>> this list
>> >> members.
>> >>
>> >> I completely forgot they existed but then this is the
>> first time I've
>> >> received a complaint about Mediafire.
>> >>
>> >>
>> >> > On Oct 26, 2016, at 5:43 AM, Roman Sowa <modular at go2.pl> wrote:
>> >> >
>> >> > Why use predatory file hosting sites while there are sites
>> >> like synthdiy.com or synth.net, which AFAIK are managed by
>> this list
>> >> members.
>> >> >
>> >> > If anyone's interested I can also host any doc files (no
>> >> audio/video/warez or any pirated stuff) at regular
>> downolad links, no
>> >> ads, no download buttons or popups.
>> >>
>> >>
>> >> Good to know. I may take you up on it.
>> >>
>> >>
>> >> Terry Bowman, KA4HJH
>> >> "The Mac Doctor"
>> >>
>> >> _______________________________________________
>> >> Synth-diy mailing list
>> >> Synth-diy at dropmix.xs4all.nl
>> >> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>> >
>> > _______________________________________________
>> > Synth-diy mailing list
>> > Synth-diy at dropmix.xs4all.nl
>> > http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>