But it’s not just for HiHats, i could observe such “NaN” symthom for some other drum modules’ outputs too, and occurs randomly after load, unfortunately.
Just noticed Hora released updates today. Unfortunately, all the updates including Analog Drums failed to load after the update (at least on my Win10 PC) so I rolled back to the previous versions. Hopefully, the update will be fixed soon.
Update (Jan 9) it looks like the mew releases were removed a few hours after release. Will post another update when fixed.
I have the same issue with RimShot at v2.2.4. This is the only version I have installed, since I just bought the module recently. What are the best options for downgrading a single module? Does anyone know the latest version that was not broken?
Need say, a bit strange circumstance with author (Raphael). I did communicate with him in december 2024 (about bug in Hora Spectral Processor, for that i made separate topic once), and he said that “sent the new plugin builds to VCV so they will likely update those soon”.
But as i see in Spectral Processor shop page info, it’s still the same 223 version as was when i sent report.
I had already reported this weird bug (who affect the commercial version) to Raphael by email (hora.music.contact@gmail.com) like you’ve wrote, March 26th to be precise!
Noboby except Raphael himself can update it, because the source code is not public (the related plugin is commercial).
Of course it’s always possible to add a kind of “voltage limiter / clipper” module before 1v/Oct input jack (in order to limit to +8.5V - or below - and avoid the Detour + VCV Rack crash, BTW), but it’s not serious and not professional I guess. However I can admit 8.5V as V/Oct is largely over audio range for humans, as example, C9 note is approx. +5.58V (mine 62-yrs old ears are not capable to hear these frequencies - lol).
I can reproduce this crash, so I figured I’d provide some extra info after testing this a bit myself:
This seems to be caused by the module not clamping the frequency to sample rate / 2, which then causes it to alias if the V/Oct is high enough. Something causes the signal to “blow up” if you go high enough for the aliased frequency to reach all the way down to 0Hz.
This makes the crash sample-rate dependant, because the V/Oct needs to be high enough for the aliasing to fold back at sample rate / 2, then reach all the way back down to 0Hz. For 48KHz, that’s about 8.6V, for 96KHz, about 9.6V. At 192KHz and above, even a full 10V isn’t enough to fold it back down to 0Hz and trigger the crash.