Vult Unstabile bug? Huge cutoff downshift at 352.8 kHz and above

Tested on Win 10 and Mac M1, Rack 2.0.6 Pro standalone

The Vult Unstable filter works consistently with a Rack sample rate between 44.1 kHz and 192 kHz

However, at 352.8 kHz and above there is a massive downshift in the cutoff frequency, making it nearly unusable at those rates.

For example, if the cutoff is ~10 kHz at 192 kHz sample rate, then bumping the sample rate to 352.8 or above drops the cutoff to ~2.4 kHz.

I did not test any sample rate below 44.1 kHz due to the following post from Leonardo:

yes, I verified this. This peak was at 1khz at 44k, 96k, 192k

Since that sample rate was not available in Rack v1 it was untested. As I mentioned in my other post, all the modules have optimized code for the common sample rates available in sound cards (44.1, 48, 88.2 and 96 kHz). For all other sample rates thereā€™s an unoptimized fallback. In the case of Stabile the fallback is incorrect.

All modules released during the Rack v1 period are not tested for the new sample rates. If you find more errors, just report them in my GitHub and I will fix them when I have time.

1 Like

I looked for your GitHub before posting here, but failed to find it. Can you share a link?

I believe Issues Ā· modlfo/VultModules Ā· GitHub is what youā€™re looking for.

Itā€™s fine to have obscure bugs, but ā€œunoptimized fallbackā€ sounds a little different from what actually happensā€¦

Let me explain.

For all common sample rates, I create precomputed tables for speeding up calculations of equations. In some cases I have specialized code for an specific sample rate. Thatā€™s what I refer as ā€œoptimizedā€ code.

The fallback functions perform the calculations for arbitrary sample rates. The bug reported is in this fallback code which was not tested. It was neither possible to hit in Rack v1 (actually v0.6) when this module was released.

4 Likes

352.8kHz seems like an absurdly high sample rate. Do you have dog ears? :dog:

1 Like

I have patched up a VCV Rack benjolin, which sends a derived highly variable pulse wave from two triangle waves and a comparator, and sends that signal to the filter input. There is no anti-aliasing on that input, so there can be significant aliasing when higher pitches are produced - the ultrasonic sound above the Nyquist frequency are reflected back into the audio range.

I was experimenting with sample rates, and heard a definite improvement going from 96 kHz to 192 kHz. So I wanted to try 384 kHz to see if I saw further improvement.

1 Like

Care to share the Benjolin?

1 Like

Absolutely - Iā€™m working on it. I still need to make a little demo video. I am loving the results!

1 Like

Iā€™ve suggested it as a module to a few devs, a couple sounded interested but no one has made one yet to my knowledge. Look forward to having a play with it.

I read somewhere (canā€™t find it now) that Rob Hordijk didnā€™t want his module in software, there was also something with one of the modules of MSM that had to be removed when the source code became open source.

afbeelding

This one had to go.

I read something he wrote, or heard him in a video (not sure which) something similar. But what I took from that was he preferred the results of hardware because it provided more subtle variation. I didnā€™t interpret the statement as meaning he would be upset if it was implemented in software. But I could be wrong.

A number of people have gotten licenses to market hardware based on his design. I suppose a VCV module would require a license agreement as well, whether free or premium. But I would be shocked if he would be upset by a VCV patch that emulates a benjolin.

2 Likes

I think a patch would be OK, but not a module.

Hi! My first post :slight_smile: I am also curious for your patch! I have made an emulation as well and will try to post it also these days. I hope to learn from your version!

I also made an emulation patch of the Wing Pinger by Meng Qi. Will post this as well. I donā€™t know if it resembles the real hardware, but the result is quite musical

2 Likes

(RANT-mode-on) Not to you, just a general commentā€¦

Iā€™m sorry but if that had to be removed thatā€™s at the level of stupid of YouTube copyright claims and takedowns. It had no Hordijk or Benjolin naming on it, looks nothing like the countless clones of Benjolins, which can be sold without problems and the Benjolin is an open, DIY, educational design anyway. It fully complies with the VCV Plugin Ethics Guidelines. So if thatā€™s true itā€™s pretty ridiculous.

(RANT-off)

Why? (see above)

Quote:

VCV Plugin Ethics Guidelines Ā¶

  • You may not clone the brand name, model name, logo, panel design, or layout of components (knobs, ports, switches, etc) of an existing hardware or software product without permission from its owner, regardless of whether these are covered under trademark/copyright law.

It is recommended to follow these guidelines for all plugins, but you are not legally obligated to do so. However, it is a requirement for:

ā† End-quote

If it doesnā€™t mention Rob Hordijk, or Benjolin, does not have itā€™s logo, does not have the same panel design or component layout, a developer should be able to make as many ā€œBenjolinā€ or other modules as they like. Yes?

Why? (see above)

(RANT-on) Iā€™m getting a bit tired of the (mostly unfounded) license and copyright paranoia in the community. Itā€™s stiffling, spreads a bad mood and is just plain annoying. Just abide by the guidelines and get on with making fun stuff. If weā€™re to become a bunch of half-assed lawyers hurling hear-say at each other, the fun will really go out the window, and has sometimes. (RANT-off)

2 Likes
1 Like

Okay, so it was the The TwinPeak Filter, not the Benjolin. Lots of clones of this being sold.

Quoting:

I made my plugin closed source a while ago because Rob Hordijk only allowed me to release the TVCF (TwinPeak filter clone) when closed source.

I highly doubt that Rob Hordijk has the power to patent or protect the ā€œprinciple of two inverse-parallel low pass filtersā€ - it smells like the dubious and illegal practice of software or algorithm patents. But I canā€™t be sure of course and Iā€™m not going to bother looking.

Michael Struggl obviously reached out to Rob who said he couldnā€™t implement the design in Open Source software (why?) and he complied. I highly doubt he has any legal bearing to do that but I canā€™t be sure and Iā€™m not a lawyer. I donā€™t see why the module should not be fine as long as it adheres to the above mentioned VCV ethical guidelines.

My take: Donā€™t ask and make sure it doesnā€™t look like an original and donā€™t mention where the inspiration comes from.

I just casually looked and Moog has three patents. The first two reads like patenting a general VCO and VCF. If people took crap like that seriously there would be no music industry, no synths or modulars and no VCV Rack. The definition of stiffling. Sighā€¦

4 Likes

@pgatt and @koen.kaptijn - The patch and demo video are now available!

2 Likes