saving a cascade~ filter...any way possible
good day,
i am designing a synthesis program and trying to save all the settings in the patch by using the autopattr object. all the objects i am using are recognized, except filtergraph (or cascade for that matter). I realize that the filtergraph and the cascade objects will not output any info for autopattr, but i need some way to save the filter settings. my patch has multiple subpachers with a filtergraph in each one, and i have an autopattr in each subpatch and one in the parent patch. also, i have tried a number of ways using the preset object within the subpatchers and using a preset object from the master patcher (with its 'included objects' cable going into the filtergraphs in the subpatch), but there is no security in this effort. preset always forgets the subpatcher's filtergraph settings after i write a file with the write command. please help if anyone knows a method or object that will allow me to complete my patch.
thanks,
kp
I can't think of an elegant way. What comes to mind first is saving the filter coefficients from filtergraph~. That would entail an [unpack] with 5 float arguments for each filter represented by filtergraph. The floats could be connected to number boxes and thus exposed to pattr. A simpler way might be to use an object like filtercoeff~. That way you only need to deal with frequency, gain and q for each filter. I recall a non-MSP object like filtercoeff~ (filter-helper or something), but I can't find it just now.
Hello,
I also save the filter coefficients (five per filter stage in a cascade), but using a multislider instead of number boxes. It remains a very boring way to do this ... I hope [filtergraph~] works with the pattr system in Max 5.
You can alternatively use [filtercoeff~] instead of [filtergraph] but in this case you need to design your own GUI. Another probem is that a 3-stage cascaded filter needs up to three [filtercoeff~].
>I recall a non-MSP object like filtercoeff~ (filter-helper or
>something), but I can't find it just now.
you might be talking about the biquad_coef abstractions made by R
Dudas - they are very handy and i use them a lot (with [pak]) to
build filters using [cascade~]
it might also be what you need
best
kasper
--
Kasper T. Toeplitz
noise, composition, bass, computer
http://www.sleazeArt.com
"you might be talking about the biquad_coef abstractions made by R
Dudas - they are very handy and i use them a lot (with [pak]) to
build filters using [cascade~]"
.................
thanks for your help!
where might i be able to find R. Dudas' biquadbank~ and what user interface do you use to input the coeffs? i would like my average user to be able to manipulate a multi-bank eq..
thanks,
kp
thanks for you contribution.
i tried settings up five filters with each one exposing its' freq, amp, q to a pack object going into cascade~, however it totally eats up CPU power, since i need to use 5 of these instances throughout my patch. i also realize my patch needs additional 'cleaning,' so i DO appreciate you ideas...
-KP
>thanks for you contribution.
>i tried settings up five filters with each one exposing its' freq,
>amp, q to a pack object going into cascade~, however it totally eats
>up CPU power, since i need to use 5 of these instances throughout my
>patch.
as said in an earlier post, I use R Dudas' small collection of
biquad_coef (with things like coef_bpf, coef_gbpf etc) - it is not
the (audio) biquadbank~
using those i build my filters (multiple filters) by outputting
those coef_objects (abstractions) to a pak, which i then connact to a
cascade~
it does eat very little cpu, really (it's just one cascade~ and some
max controls - having much more than 5 instances is not a problem (I
still use a PPC G4, no problem at all)
I belive i found those on C74 web site, in "share" or something like this
best
kasper
--
Kasper T. Toeplitz
noise, composition, bass, computer
http://www.sleazeArt.com
Kasper T Toeplitz schrieb:
> I belive i found those on C74 web site, in "share" or something like this
could you elaborate on this? they are neither on R. Dudas share site nor
in Maxobjects. The closest I could find seems to be the ISPW
compatibility library, but that's only for IRCAM members, and the object
names seem to be different so I can't verify...
Stefan
--
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com
i have been searching for these objects as well...to no avail.
>Kasper T Toeplitz schrieb:
>>I belive i found those on C74 web site, in "share" or something like this
>
>could you elaborate on this? they are neither on R. Dudas share site
>nor in Maxobjects. The closest I could find seems to be the ISPW
>compatibility library, but that's only for IRCAM members, and the
>object names seem to be different so I can't verify...
>
i don't remember where i found those - I think it was on the "share"
page of C74...
it is NOT ispw for sure...
the package is named "biquad_coef", there are 11 elements in it
(coef_bpf, coef_gbpf, coef_ghpf etc) and help files as well - I
really belive/remember it was done by R dudas, and the objects are
actually abstraction, max only, which take 3 inputs (typically freq,
amp, q) and output "something" to input in the 2nd infrom left on a
biquad~
I use those, with a pak, to build more complex filters, outputting
the pak into right in of cascade~
suppose (??) it was a free distribution, so i could put them
somewhere for download... or not??
best
kasper
--
Kasper T. Toeplitz
noise, composition, bass, computer
http://www.sleazeArt.com
On 6 nov. 07, at 16:00, Kasper T Toeplitz wrote:
>> i don't remember where i found those - I think it was on the
>> "share" page of C74...
>
> it is NOT ispw for sure...
Richard told me "they are (in an early version) somewhere in the
cnmat nightschool collection of patches from the late 90s." I'm sure
Michael is gonna confirmed that ;-)
ej
Emmanuel Jourdan schrieb:
> Richard told me "they are (in an early version) somewhere in the cnmat
> nightschool collection of patches from the late 90s." I'm sure Michael
> is gonna confirmed that ;-)
Yes, If found them, CNMAT Nightschool 1999, Dudas Goodies...
Seems I didn't do the tutorials... ;-)
Stefan
--
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com
Quote: Stefan Tiedje wrote on Thu, 08 November 2007 09:02
>
> Yes, If found them, CNMAT Nightschool 1999, Dudas Goodies...
Could you share a link? I can't seem to find the night shool archives.
Regards,
kjg
Hi,
I actually could not find that night school session in their archives. I looked everywhere and asked others in this post if they had the objects with no success. What i ended up doing for my patch was using a series of filtergraphs (next to each other) as a sort of limited multiband eq. obviously this might not always be ideal, but it worked for that project. if you want me to send you that example i'd be glad...just let me know.
-kp
>Hi,
>I actually could not find that night school session in their
>archives. I looked everywhere and asked others in this post if they
>had the objects with no success. What i ended up doing for my patch
>was using a series of filtergraphs (next to each other) as a sort of
>limited multiband eq. obviously this might not always be ideal, but
>it worked for that project. if you want me to send you that example
>i'd be glad...just let me know.
>-kp
>--
sorry it took me long to answer your mail
yes I can send you an exemple, and yes i can send you the objects (I
hope it is not "wrong" !!) - probably not on the list.... but some
private email, or something
best
kasper
--
Kasper T. Toeplitz
noise, composition, bass, computer
http://www.sleazeArt.com