Hello Claude,

thanks, i did not knew the reference 2. i did not find in any of them a low pass resonant filter but i'll have to investigate.

thanks, cyrille

Claude Heiland-Allen a écrit :

cyrille henry wrote:

...

or anyone have references in order to disign filter?

http://crca.ucsd.edu/~msp/techniques/latest/book-html/node139.html http://www-users.cs.york.ac.uk/~fisher/mkfilter/ http://ccrma.stanford.edu/~jos/filters/

Hope these help,

Claude

Hello Claude,

this is very nice.

changing resonnance does affect the amplitude, so it's quite hard to use this parameter. so i think it's possible to make a better one, but i'll use it for now.

thanks Cyrille

Claude Heiland-Allen a écrit :

cyrille henry wrote:

...

i did not find in any of them a low pass resonant filter but i'll have to investigate.

I made a (non-scientific, but sounds ok to my (bad) ears...) resonant low-pass filter, see attached. two zeros at 1/2 sampling rate, plus two poles (conjugate pair) that can be moved about controlled by audio input. The gain formula comes from Miller's book, iirc, and I probably misused it, but it seems to work ok.

...

thanks, cyrille

Thanks,

Claude

---

---

PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list

yeah roman, that peak filter is lush.

has anyone else noticed that the rpole~ / cpole~ filters use quite a bit of cpu?

hello Roman,

it's a nice filter.

here is a butterworth low pass filter that came with exemple H13. i just change expr to expr~ to allow an audio control of the cutoff frequency.

cyrille

Roman Haefeli a écrit :

hi cyrille

i am glad to hear that someone has similar thoughts. i would love to have some filter abstractions based on pd's built-in elementary filters. however, my understanding is far too limited to create them myself.

i once ported and adapted a peakfilter from millers examples into pdmtl abs (are they now part of pd-extended, actually?). it is called

[fx.filter.peak.sig~]

the sig stands for audio frequency inlet and audio q inlet. especially modulating the q parameter with a signal is fun.

to hear it in action (besides a lot of other modulations): http://www.netpd.org/sessions/2007-08-17_nordatlantik.mp3

roman

On Fri, 2007-11-23 at 20:54 +0100, cyrille henry wrote:

...

hello,

i know there is a lot's of filter externals available for pd, but i'd like to use some using rpole~ / cpole~ etc. my main motiviations are :

1. i don't like using externals
2. this object allow an audio control of the cutoff frequency

as there is not a lot's of exemples, i'd like to know if anyone used this objects to make good filter abstraction. or anyone have references in order to disign filter? (i'm looking for a low pass resonant filter by exemple)

thanks Cyrille

---

PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list

___________________________________________________________ Der frühe Vogel fängt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de

---

---

PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list

• Chuckk Hubbard badmuthahubbard@gmail.com [2007-11-24 09:28]:

On Nov 23, 2007 9:54 PM, cyrille henry cyrille.henry@la-kitchen.fr wrote:

...

hello,

i know there is a lot's of filter externals available for pd, but i'd like to use some using rpole~ / cpole~ etc.

I like playing with raw filters too. I don't have any references to add to Claude's, and really for me it's more a matter of getting through the references I already have. Miller's book has a very good explanation, lots of graphs, lots of explanations for the rotation and complex number stuff, but still lots of detail. I've also used the graphing abstractions in H14.all.pass.pd, substituting rpole~ et al to see the frequency and phase response as I change the values. I also found the [fexpr~] object educational. I wonder what more knowledgeable people will say to this: my impression has been that, as interesting as filter design is, and although it's possible and not terribly difficult to learn to understand it well, it is not something that will necessarily become intuitive with practice. That, to design specific filters, it will still be a matter of performing series of calculations to know what values to use, rather than "sort of knowing" what to put. Thoughts?

Yes, master the Z-Transform (which is not hard, really). :-)

regards, PP

-Chuckk

-- http://www.badmuthahubbard.com

---

PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list

Charles Henry wrote:

I remember there were many filter abstractions based on biquad~s. I thought they were in the iemabs directory, but I can't find them.

there are abstractions in ggext that do this.

fgmad.r IOhannes

cyrille henry wrote:

hello Ioannes

IOhannes m zmölnig a écrit :

...

Charles Henry wrote:

...

I remember there were many filter abstractions based on biquad~s. I thought they were in the iemabs directory, but I can't find them.

there are abstractions in ggext that do this.

you mean ggee?

yes

there is only externals there.

true (i checked now :-))

or did i miss something?

i don't think so. it was my fault. but i guess most of ggee's filter externals can be rewritten as abstractions (i checked only [highshelf] and i haven't seen anything you cannot do with plain pd.) the only problems that might occur are about precision (as you can only use single precision in abstractions), but at least highshelf.c uses single precision too, so that should be ok.

fgmasd.r IOhannes