How do some filters allow the cutoff/center freq to be above SR/2?

[Nitsuj70]

Very nice!

The original link has moved…to here: ZLEqualizer/source/dsp/iir_filter/coeff/martin_coeff.hpp at main · ZL-Audio/ZLEqualizer · GitHub

I’m guessing that the usual issues with (RBJ style) biquads and direct-form are mitigated because of using doubles? Things like blowing up under fast modulation etc.

Coming back to the original question, do these filters allow cutoff to be above Nyqust?

[zsliu98]

Thanks for updating the link

@Nitsuj70

I’m guessing that the usual issues with (RBJ style) biquads and direct-form are mitigated because of using doubles? Things like blowing up under fast modulation etc.

AFAIK the double precision only makes the filter stable in low-frequency range. If you want to do fast modulation (and don’t care much about phase), you can use SVF instead of TDF-II. The SVF coefficients can be calculated from biquad coefficients:

dafx14_aaron_wishnick_time_varying_filters_for_.pdf

ZL-Audio/ZLEqualizer/blob/main/source/dsp/iir_filter/svf_base.hpp#L98

@Nitsuj70

Coming back to the original question, do these filters allow cutoff to be above Nyqust?

The matching method should work. But you have to pick the matched points very carefully, especially for shelving filters.