Dear list,
I am writing a patch for FIR filter design in the old way, by realizing the formula y(n)=a(n)+b(n-1)+c(n-2)+...+?(n-??) with z~. All the work is ready, but I am just missing the designs. I wanted to make also some presets and the ability to design the filter from normal parameters (type, cutoff freq, etc). Does anybody knows how to make the calculus for FIR filters, to know the equation?
By the way, there is a bug in the dynamic instructions: when the instruction "vis 0" is given to a already closed subpatch, the program crashes.
Thanks as usual,
João Miguel Pais
hi João,
there used to be a crash-case of sending a 'find' message to a closed window, corrected in one of .35 test versions. Not sure if the fix has been equally enough for linux and msw.
Just sending the 'vis 0' (without the following 'find', 'cut', etc.) is not known to produce crashes, at least it is not reproducible here.
Could you please describe your case better, like including info about an os, pd version, and, possibly, a context?
Krzysztof
btw, the normal' way of applying a fir filter would be to read ir from an array by a specialized object like buffir~ of msp (not yet in sickle, sorry:). But, anyway, for modeling thenormal' filter
parameters, I would suggest designing iir filters with biquad~.
João M Pais wrote: ...
I am writing a patch for FIR filter design in the old way, by realizing the formula y(n)=a(n)+b(n-1)+c(n-2)+...+?(n-??) with z~. All the work is ready, but I am just missing the designs. I wanted to make also some presets and the ability to design the filter from normal parameters (type, cutoff freq, etc). Does anybody knows how to make the calculus for FIR filters, to know the equation?
By the way, there is a bug in the dynamic instructions: when the instruction "vis 0" is given to a already closed subpatch, the program crashes.
I am using .35 t26 on xp. In my patch I create a window with the objects, and I can also kill it, with the message "editmode1, selectall, cut, editmode 0, vis 0". But if this message is banged when the window is already closed, the program crashes.
Regarding the "find last error" bug, I tried it and if there wasn't any error before, it runs fine (says that there was no error before). But when there are some mistakes (I tried connecting the output from an osc~ to a slider), the program still crashes.
João Miguel Pais
----- Original Message ----- From: "Krzysztof Czaja" czaja@chopin.edu.pl To: "João M Pais" jmmmpl@netcabo.pt Cc: "PD" pd-list@iem.kug.ac.at Sent: Saturday, August 24, 2002 2:52 PM Subject: Re: [PD] FIR filter design
hi João,
there used to be a crash-case of sending a 'find' message to a closed window, corrected in one of .35 test versions. Not sure if the fix has been equally enough for linux and msw.
Just sending the 'vis 0' (without the following 'find', 'cut', etc.) is not known to produce crashes, at least it is not reproducible here.
Could you please describe your case better, like including info about an os, pd version, and, possibly, a context?
Krzysztof
btw, the normal' way of applying a fir filter would be to read ir from an array by a specialized object like buffir~ of msp (not yet in sickle, sorry:). But, anyway, for modeling thenormal' filter
parameters, I would suggest designing iir filters with biquad~.
João M Pais wrote: ...
I am writing a patch for FIR filter design in the old way, by realizing
the
formula y(n)=a(n)+b(n-1)+c(n-2)+...+?(n-??) with z~. All the work is
ready,
but I am just missing the designs. I wanted to make also some presets and the ability to design the filter from normal parameters (type, cutoff
freq,
etc). Does anybody knows how to make the calculus for FIR filters, to
know
the equation?
By the way, there is a bug in the dynamic instructions: when the
instruction
"vis 0" is given to a already closed subpatch, the program crashes.
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
hi João,
João M Pais wrote:
I am using .35 t26 on xp. In my patch I create a window with the objects, and I can also kill it, with the message "editmode1, selectall, cut, editmode 0, vis 0". But if this message is banged when the window is already closed, the program crashes.
the crashes are caused by the 'cut', not by the 'vis 0'. This is a known problem, reported on pd-list before -- for the time being simply do not cut in a closed window.
Btw, if there was no crash, the 'cut' in a closed window would simply do nothing.
Regarding the "find last error" bug, I tried it and if there wasn't any error before, it runs fine (says that there was no error before). But when there are some mistakes (I tried connecting the output from an osc~ to a slider), the program still crashes.
This one (windows-specific) has been corrected in pd-0.35-1.
Krzysztof
Joco
did you sort the equation out? - I have just read your email and am confused about your equation - do you mean y[n] = ax[n] + bx[n-1] + cx[n-2] + .... x[n] being the digital input to the FIR (x[n-1] being the input delayed by 1 time sample) and a,b,c being the coefficients of the impulse response of the filter that you require, the number of coefficients therefore dictating the final x[n-?] delay that is required to implement the FIR - ie convolution of input and FIR? I presume to dictate cut-off freq of the FIR you are going to have to work in the freq domain as opposed to the time domain - or at least for the calculation- or am I getting very confused here?? just interested
Mandy
On Sat, 24 Aug 2002 01:04:00 +0100 João_M_Pais jmmmpl@netcabo.pt wrote:
Dear list,
I am writing a patch for FIR filter design in the old way, by realizing the formula y(n)=a(n)+b(n-1)+c(n-2)+...+?(n-??) with z~. All the work is ready, but I am just missing the designs. I wanted to make also some presets and the ability to design the filter from normal parameters (type, cutoff freq, etc). Does anybody knows how to make the calculus for FIR filters, to know the equation?
By the way, there is a bug in the dynamic instructions: when the instruction "vis 0" is given to a already closed subpatch, the program crashes.
Thanks as usual,
Joco Miguel Pais
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
MLN Kazem, Aerospace Engineering M.Kazem@bristol.ac.uk Tel. 0117 9288219 Fax. 0117 9272771
-
João M Pais -
Krzysztof Czaja -
MLN Kazem, Aerospace Engineering