yes, I had seen that graphic as well. but trying out the highpass
iemfilters with noise, turns out that 10 hip~s work better.
Hehe, I guess that's because you filter noise: It's harder to recognize
the
noise that the 10 hip stages added. :)
that's true. I started then using other signals with varying complexity,
and the output was better. btw, I was testing the signal only by looking
at it, through a spectral analysis.
I looked in my list of pd-ext objects, but found no elliptic filters.
are
there any, or any way of getting their coefficients to use with biquad~?
Elliptic filters are rather hard to design, most of the time you will
find
Butterworth or Chebyshev designs in our little audio world.
A biquad~ has only two poles and zeroes, so you would need to cascade
several
stages to get a really sharp filter.
yep, I only found those. I'll try cascading other types of filters as
well. I thought there would be a magic filter somewhere that would do any
operation imaginable by men.
_______________________________________________
Pd-list@iem.at mailing list
UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list
