Yeah, I was just plugging Vult in a way that was off topic. Cytomic is pretty great too!
I understand that the Vult modules for a long time were pretty much the only option for quality (edit: quality in terms of highly accurate emulations of classic non-linear resonant filter circuits that are very popular in both Eurorack and standalone synths, for example the MS20 mk1 / mk2, Jupiter 8 / SH-09, Prophet, Oscar, Moog ladders, etc etc, to the point you can compare both the model and the circuit side by sound and they exhibit extremely close behaviour at all settings of drive, cutoff, and resonance etc so the waveform and the spectrum match very closely), so they have a strong and well deserved following.
Donāt worry, you can make up for it by going to any Ferox threads and post how good the A-124 module is
Ummā¦ Only option for quality. ok, sparky, if it makes you feel good to think that, thatās just fine.
of course there is no doubt about the quality of vult, but they are not the only modules with high quality, starting with the fundamental VCF there is a long list of good filters and other modules in the VCV Rack, we will be able to welcome to the wasp module to that long list
Thanks for letting me know that there are some other good options. I formed my opinion because there is no way to trial modules, so had to rely on asking a bunch of existing users of VCV Rack for options as to for the most detailed non-linear filters in VCV Rack, they all said Vult. I was literally not even told of a single other module maker to check out, so input in this area would be appreciated.
As for the VCF module, it is a vanilla textbook Moog ladder emulation using RK4 (4th order Runge-Kutta) so is limited to 8khz cutoff frequency (for 44.1khz sample rate). The high pass doesnāt work very well either, I did help Andrew a bit, but there is no helping RK4 without oversampling x2 or x4 - itās an explicit method so just canāt resolve high frequencies very well. The module is useful sure, and great to have included in fundamental, but itās not really what I think of as a detailed analog filter model.
you know @andy-cytomic , I m animator, I m very far to be a filter expert, in my work, exact simulation is not synonymous with qualityā¦
Ok, so that may be the issue. The quality I was referring to in this context meant detailed and accurate highly non-linear analog resonant filter models, sorry if there was confusion. Please do still point out any VCV Rack filters, apart from Vult, that I can test against my collection of analog filters: MS20 mk1 and mk2, Moog Ladders of various kinds, SEM, X-pander, Dark Energy mk1 (SSM cascade), any of the classics would be fine.
True for synth modules, too.
That was very far from obvious. And now true, there are not that many detailed analog models. There are some crazy people who do not believe the two are synonymous, quality and replica of an existing old analog circuit.
@andy-cytomic - at some stage Iād love to have a crack at properly modelling the Befaco BF-22 which is MS-20 inspired. Do you have a particular paper/papers that youād reccomend I take a look at (of yours or otherwise)? I do have a background in physics/computational modelling, but have yet to really apply that in a DSP setting.
The BF-22 is a straight up MS-20 mk1, but I like they added VC control of the resonance. Itās quite a difficult circuit to tackle for someone new to circuit modelling, you have to model non-linear elements like diodes, bjts, j-fets, op-amps, and otas, and then all the regular passive components like resistors and capacitors. Perhaps pick some circuits with just one of the non-linear elements to get that working first, then build up one by one.
I did an introductory talk at ADC 2020 (and available on youtube) called āFrom Circuit To Code: Under the Hood of Analog Modellingā which will hopefully help, and at the end of the included slides are lots of references for further reading: Technical Papers ā Cytomic
Thanks for the information! And yes I agree about building up slowly, itās defintely more of a long term goal. Iāll take a look at that talk too, thanks.
The interesting thing to me is that most MS-20 filter clones are clones of the later OTA filter, but the A106-1 is supposedly based on the original Korg 35 filter that was in the MS-10 and early MS-20 models so isnāt the same as all the other MS-20 filter clones out there.
Yep, Iām pretty sure with all the noise complaints Iāve read on forums that the A-106-1 is based on the MS20 mk1 filter with the Korg 35 module. They are noisy, but very fun, and sound very different the the mk2 versions. I like both of them really, and even more variations on the basic Sallen Key structure.
Edit: I see now that the comment was clarified further down, rendering this post fairly needless. The point had already been made, and hammering it down further seems is a bit meaningless and cruel, and wholly beside the point of this thread. As such, Iāve opted to removed my comment. I apologise for any discomfort caused.
The word āqualityā I think has hit a nerve, so please feel free to substitute āaccurate and detailed analog model of real hardware that matches a popular Eurorack highly non-linear resonant filterā, but that was a little long to write so I was just going with the context of this thread, which is about an accurate emulation of a much loved highly non-linear resonant filter circuit. I will edit my previous message to clarify this so people donāt get so worked up coming to this thread late.
Itās great you feel that doing accurate analog modeling to match particular hardware isnāt needed, and I agree that you can make great sounding modules without matching analog hardware, and Iām all for new stuff. I also feel that there are many great sounding and popular Eurorack modules, and they are popular for a reason, their sound as well as their functionality, and I think these shouldnāt be ignored.
@3HR You posted a bunch of modules before the edit. Are there any that actually model a bit of Eurorack hardware accurately? If so please post me the names of modules, and which original hardware they model as Iād love to check them out. Likewise please feel free to compare the A-124 model Iāve done with the hardware and let me know what you think
You can actually filter on the tag āhardware cloneā in the library, which is a far easier way to check that out rather than me listing everything by hand.
That said: I suspect that most of them are clones of modules that are digitally based, which I suppose isnāt quite relevant to your question. Furthermore, most of the hardware emulations that are analog-based are released by the companies who originally developed them. I think Alright Devicesās T-Wrex and Erica Synths Fusion Delay are good examples of this.
As for the A-124, I donāt have the hardware myself to make a comparison. Iām a purely digitally based developer, so I donāt even have the tools to make the comparison even if I had the hardware module. However, your video speaks for itself. The only thing I noticed was that the knob positions for the cutoff seem to not be 100% accurate, with the VCVRack module often needing to be one āstepā below the eurorack module for them to match, though this seems to be a design decision as it looks like the module is able to achieve a higher cutoff-value than the eurorack module?
All that said: it sounds great, and the accuracy is an impressive feat.
That is correct. The original circuit reaches 12 kHz, but we didnāt think that was enough since most of our ears can detect frequencies higher than this, so the circuit model supports cutoff up to 30 kHz, and therefore the Frq. knob has more range.
Haha, going from 12k to 30k is an almost comical extension of the frequency range, if the motivation was to include a larger part of the hearable spectrum. My cats will thank you for the heightened fidelity in the high-end.