I checked the self oscillation output frequencies, and I am finding the tracking to be better than what you report by an order of magnitude. The error reported by the Nysthi Hot Tuna varies a bit by VCV sample rate, as well as by frequency. But at 48 kHz going from C4 to c8 I find that setting the attenuator to 99.96% gives a “perfect” result.
More importantly, I checked the self oscillation output against the Venom VCO Unit sine as well as the VCV VCO sine, and all were in perfect agreement with each other. So I am wondering if we are running up against the limits of floating point math and sample rate.
So I will not make any tuning adjustments to the filter.
Not criticizing. Just noticing. It can follow V/Oct, that’s the main takeaway. So tuning differences would not occur due to using different sample rates, because I was also running 48 kHz.
But, I did not use a tuner to check, because a tuner might have difficulty tuning to resonant (but still narrow band) noise. I just checked by ear (and minimizing the detuning beating) against two sine and V/Oct sources. In this case using two instances of BogaudioREFTONE which let you easily set/select notes and octaves, and conveniently outputs both a V/Oct and a Sine. So, could just be my aging and failing auditory perception against the accuracy of a tuner. Definitely don’t fix what ain’t broken.
Whoops, shame on me. I did not read any of the documentation before fiddling knobs.
Finally, on using the Multimode Filter(s) as resonator(s).
Feeding it variations of (low pass) filtered noise yields great varieties of distortion behaviour in sweet spots. E.g using Squinky LabsColors.
Again, another great module. Thanks for your evergrowing contributions to the VCV Rack community.
The Benjolin patch is one of the most outstanding patches I’ve ever had the pleasure of using. Thank you so very much for your amazing dedication and hard work.
I have submitted v2.14.2 to the library. I don’t know when the update will appear in the library. In the mean time you can get the binaries below. EDIT 2025-12-07 - v2.14.2 is now in the library
Version 2.14.2 Changes
Enhancements
Multimode Filter
Five new Morph modes
BP ↔ Notch
Dry ↔ Wet LP
Dry ↔ Wet HP
Dry ↔ Wet BP
Dry ↔ Wet Notch
New input coupling option button
DC (default) - old behavior
AC - This can eliminate saturation asymmetry when the input has a DC offset
New Gain VCA polarity option button
Unipolar (default) - old behavior
Bipolar - This can only have an effect with Gain CV
Gain range extended to 10
CV now scaled at 1 per Volt instead of 0.2 per Volt
Effective gain now clamped to 0V - 10V or -10V - 10V, depending on VCA polarity
This can be a breaking change for old patches with Gain CV
Wincomp
Added context menu option to normal B input to the previous sample from the A input so Wincomp can function as a crude slope detector.
Bug Fix
Multimode Filter Notch output was sometimes inverted
The filter is great! Made a nice ping patch with it the other day. Appreciate the super steep slopes too, those are kinda rare in VCV and can sound great sometimes.
Version v2.14.3 | v2.14.4 v2.14.5 has been submitted to the library with a fix for a long standing bug in Linear Beats as per issue 33, reported right in the middle of the last release to the library.
In addition there is a fix for XM-OP sometimes crashing on patch load if quantize is disabled as per issue 34. Thanks to discord user phantombeta for localizing the bug, making it quick and easy for me to fix it.
Multimode Filter enhancement
The Spread Direction now has a new Right Absolute option that causes the Cutoff knob and CV to only apply to the left channel, and the Spread becomes the right cutoff.
v2.14.5 Binaries
If you want the fix and/or enhancement before it hits the library, you can get it below.
Another bug (this time for XM-OP) was reported while waiting for 2.14.4 to hit the library (see previous post). So the version was bumped to 2.14.5, and the library submission was updated.
Versions 2.14.3 and 2.14.4 will never make it to the library - they are superseded by 2.14.5.