There was a thread a while back that addressed this I think, but I cant find it... about the click artifacts when changing the frequency on the filter~ & vcf_filter~ based iemlib objects. It's a shame since these seem to be the most dynamic filters available, but only seem useful when static. Has anyone found a way around this?

Here is a patch comparing vcf~ to vcf_lp4~:

vcf-test.pd

@katz Hello Katz....... I cannot reproduce your problem. Does bumping up the audio buffer (delay-ms in audio settings) a notch or two make any difference?

W-A,
No difference. This is what I'm hearing:
vcf_lp4~.mp3.
What version are you using? I tried this with 0.42.5 & 0.43.4 (OSX & Windows).

@katz 0.43.4extended on win7.........
I don't seem to be able to access your .mp3 upload...........
David.

This patch fixes the issue to an extent. There were two different unwanted sounds that I was hearing: 1) a high frequency tambourine-like sound. This went away if I put the filter in a subwindow and upsampled 2x. 2) a pop sound when a large change in filter frequency was made quickly. I created a dual-filter crossfade for this. Each has a switch~ to cut down on CPU, but the vcf_filter~ type which originally used less CPU than the vcf~ now uses more.

Sounds like maybe this isn't an issue for some people, but here it is anyway:

vcf-crossfade.zip

(mp3 with the high frequency sound is included here so it doesn't have to be played in the browser)

@katz strange..... I don't have those clicks at all......??....... and they are pretty horrible. Strange though because they sound much like a cut to a new frequency waveform that is not at the zero crossing. I would expect that to happen unless it was already allowed for in [vcf_filter] and I wonder why I don't have them?
I like the x-fade patch... It takes me back to the 1970's......"lost in space"..... "the twilight zone".... all that!
Nice one!
David.

@whale-av
Thanks for the info - I'll continue to look into it. The upsample isn't even necessary for lowpass because vcf~ behaves similarly to vcf_lp2~ (some differences with the way the q and amplitude behave) & if you want vcf_lp4~, just stack two vcf~s etc etc. For right now it's useful for the bandpass and highpass and the other "filter~" filters.