On 13 Feb 2009, at 14:22, carlo-laurenzi wrote: > > Hi, does anyone know if an Msp external for calculating variable > order Chebychev filters has ever been implemented? > Thank you very much. >

i have one (osx) sitting on my hd, waiting to get finished. but the filter is pretty unstable for high orders (still makes cool noises just before it explodes!) if you like to test it, please contact me off list.

here is one that has similar problems as mine:

v

Quote: ACSmith wrote on Sun, 15 February 2009 10:50 ---------------------------------------------------- > Check out the Percolate externals by Brad Garton etc.--I believe it's > Gen14 (maybe Gen17) that implements the Cheby filter.

i think you are mixing things up. the original question was not about waveshaping (using chebyshev polynomials for generating transfer functions) but about filtering (using chebyshev polynomials for calculating filter coefficients). vb

Hi,

we have plans of implementing it in TTblue, so that it might be available in the Tap.tools tap.filter~ object and the tt.filter~ object shipped with Jamoma, but unfortunately we haven't gotten around to it yet.

Butterworth 1, 2, 3 and 4th order filters can be found in tap.filter~ and tt.filter~ though.

Best, Trond

On Feb 13, 2009, at 2:22 PM, carlo-laurenzi wrote:

> > Hi, does anyone know if an Msp external for calculating variable > order Chebychev filters has ever been implemented? > Thank you very much. > > Carlo >

Not sure it's what you are looking for, but if you own Jitter, have you looked into [jit.bfg]?

Hi Carlo - Not sure if this will help you or not, but you can always re-express any filter as a series of biquad~ objects. So if you can do the math by hand (or using matlab, or whatever) you can implement the filter in MSP without needing to write a custom external.

Or were you talking about doing the coefficient calculations?

Ben

On Fri, Feb 13, 2009 at 5:22 AM, carlo-laurenzi wrote:

> > Hi, does anyone know if an Msp external for calculating variable order > Chebychev filters has ever been implemented? > Thank you very much. > > Carlo >

Hi. yes, I was talking about doing the coefficient calculations. best,

Carlo

for anyone still interested in this thread: i've been working a little on the stability of the filter and it seems to be ok now. you can find an osx version here:http://www.esbasel.ch/Downloads/MaxMSP-Objects.htm

for the sake of crossplatform support, there is also an mxj java class, which calculates the coefficients only. use it with [cascade~] to do the actual filtering (vb.chebycoeff).

carlo, if you are still listening, i would suggest you try this new version over the old one i've send to you.

both externals let you set the number of poles (steepness of roll off), the amount of ripple in the passband (0% would be a butterworth design) and the filter mode (lp/hp).

vb