[Csnd] filter2
| Date | 2013-01-04 15:57 |
| From | Anders Genell |
| Subject | [Csnd] filter2 |
Hi!
Some of the following should maybe go into the dev list, but as most dev read this list too, I suppose it will end up in the right hands anyway.
I have a question and a comment about the use of the filter2 opcode. Comment first: In the manual it says the syntax is ares filter2 asig, iM, iN, ib0, ib1, ..., ibM, ia1, ia2, ..., iaNbut the actual syntax is a-coeffs first and b-coeffs second, which is also the way it is used in the included example. Question:
The manual also says coeffs could be introduced from a separate file e.g. using GEN23. I tried creating a table and just give the table number as arg to filter2, but that did not work. Does anyone know how to get values out of a table and into filter2?
Incidentally, filter2 is one of the opcodes on Menno's 'unfinished' list, so no shadow should fall upon his enormous effort!! Maybe we could write up an extended example after having solved these issues?
Best regards, Anders |
| Date | 2013-01-04 18:04 |
| From | menno |
| Subject | [Csnd] Re: filter2 |
Hi genell, filter2 has an example: http://www.csounds.com/manual/html/filter2.html but it is not a very good one; it does not explain a lot about and probably i have found some values empirically. So it would be great to have a better example here, explaining different settings. zfilter2 is on the unfinished list however, desperately awaiting a lovely example. Anyone for filter2 and zfilter2? -- View this message in context: http://csound.1045644.n5.nabble.com/filter2-tp5719182p5719184.html Sent from the Csound - General mailing list archive at Nabble.com. |
| Date | 2013-01-04 18:18 |
| From | Anders Genell |
| Subject | Re: [Cs-dev] [Csnd] Re: filter2 |
| Attachments | None None |
Ah, yes, my mistake! [I'll post this to dev list, since this is more related to that side, I suppose] I suppose zfilter2 works just as filter2, but has k-rate coefficients for moving poles around in real time?
I found that at least filter2 also ha a limit of 51 b-coeffs and 50 a-coeffs (the first a-coeff is assumed to be 1). Is there any reason to keep this limit for filter2? Since all coeffs are i-rate they are not time critical as such, as they are just used during initialization. It would be nice to be able to do some matlab ar-modeling of higher orders, say 256 or even higher, and just implement into csound using GEN23 and filter2....
Regards, /Anders On Fri, Jan 4, 2013 at 7:04 PM, menno <nabob_cd@yahoo.com> wrote: Hi genell, |
| Date | 2013-01-04 18:26 |
| From | Anders Genell |
| Subject | Re: [Cs-dev] [Csnd] Re: filter2 |
| Attachments | None None |
Thinking a litter further, it would be a nice feature, if the GEN23 way would work, to use the table morphing opcodes and zfilter2 to morph between different filters... just a thought.
/A On Fri, Jan 4, 2013 at 7:18 PM, Anders Genell <anders.genell@gmail.com> wrote:
|