Hardware Physical Modelling


I bought one of the new Ladik LFOs called Delayed LFO. Seems to be ideal for adding either fade-in vibrato effects (which brass players incorporate) or a fade-out burst at audio rate that gets triggered at the initial part of the sound. Unfortunately it only goes up to 30Hz. According to the SOS article the ideal burst for brass is at 80Hz. I’m waiting for mine to arrive so I’ll report back with my findings. I already contacted Ladik and he said it’s possible to triple the frequency range but the internal filtering of the module also needs to change, which will lead to less edge/more rounded waveforms.
I decided against the modification as I aim to use a triangle LFO for vibrato at the end of a long sustained note; for brass.

Might be something of interest for you.

I still haven’t tried the DC offset. Has anyone utilized inverse feedback in their patches? I’m considering swapping my Serge Resonant EQ for the Random*Source version in order to save space and take advantage of the internal feedback option. I’m curious how inverse feedback will sound - I assume it’ll hollow out the sound, which could be interesting.

Also, check the Elby Velocitizer module out. I recently found out about it and it seems interesting as an enhancer of breath control.


Hollowing is a pretty good assumption of how an inverted feedback path will change the sound. Much rounder, more generally suited for tube and membrane type sounds. Whereas positive feedback is what I’d use for strings or more clangorous percussion. Mixing the two using two separate delays can be really great too, mainly for percussion. Try setting up one delay with positive feedback, the other with negative, and feeding them through each other.

The other cool thing about inverting the feedback path, at least what I’ve found, is that it will give you a separate note that harmonizes with what you get from a positive path (not sure what the exact interval is). So patching your feedback through a ring mod or any sort of voltage controlled inverter, and using another signal to swap back and forth, can be a great way to play around with varying a sequence.


I assumed that feedback > inverter would also work the same way. I’ll give it a shot that way as well. I like your approach: try everything and see what works! :slight_smile:


that ladik lfo look interesting indeed!
i always test how the inverted feedback sounds, sometimes the effect isn’t that drastic, sometimes, the whole patch changes…for example the serge wave multiplier is an insane feedback module, especially the lowest section, and inverting the signal changes the sound quite a bit:

(not that fitting for this physical modelling thread but still interesting for feedback patches, all serge modules haha :slight_smile: so yeah, good choice to get the r*s resonant eq @ParanormalPatroler!)

with the d0 i found that inverting the feedback also gives me a more stable feedback when using the module as a bread&butter delay, dunno why this is, but the feedback is more stable and doesn’t get out of control that fast…
i have yet to try mixing positive and negative feedback, good idead @smbols!


Has anybody been able to patch up a decent string sound? I’ve tried over and over and I just find the sound of physically modeled strings to be really lackluster, outside of the Kaivo VST which produces lovely string sounds, but that’s a good bit more intricate a system than I feel would be feasible with a modular synth. Even every string sound I’ve heard people make with Rings, it all just sound like a cheap keyboard preset.


@smbols I don’t even try anymore, although I know I should give it a shot. Software is so much better in that regard - I’ve heard some amazing results here and there. I’m also happy with the results I’m getting from VL70m’s strings. As always, it’s all about how the actual sounds are being played.

Do check the Velocitizer module I posted about earlier. It’s supposed to be based around envelopes from string players. Maybe that would help?


I’ve studied and researched making string sounds for about 5 years now, exploring many different approaches and techniques.

My conclusion is more or less:

You can synthesise highly musical sounds that contain some of the characteristics of say, a violin. But to reproduce the expression - the morphology of notes or phrases, the resonances of the body and strings, the harmonic movement, the distortion, the noise, the elasticity - in a way that consistently feels like an actual string instrument is to all intents and purposes impossible. Being in a relationship with a phenomenal violinist has confirmed that to me.

But, like I said, exploring this line of sonic enquiry can absolutely yield wonderful, beautiful and strange sounds


I’ve had decent results with a mangrove and mungo c0. Will post ASAP.


Do any of your approaches and techniques involve a modular system? If so, would you care to give some pointers in the ones that have given you the best results?


I work between a computer and a modular synth most of the time, creating raw sound in the modular for use in the computer, or using the modular to send midi to the computer, or both together in a kind of spider web configuration. I never work solely on the modular though.

one approach I’ve explored has been using oscillators, feedback, karplus delay lines and filters to create raw waves that are full of harmonics, and waves of resonance that move unpredictability, rising or falling in intensity according to the amount of feedback. A drone, basically.

After recording large chunks of this sound, i would then open up five kontakt samplers, with the SIPS expression code running, all with slightly different expression parameters - different vibrato, different attack, sustain and release times, glide times, and so on. I also have on here five separate channels instances of a noise reduction plugin, usually izotope, also set to different configurations - essentially allowing more or less harmonics through, as well as changing the noise / tonal balance.

I would then send five channels from just friends, or tides, to the respective volumes of the channels, as well as the parameters on the noise reduction plugins. This would be a basic starting point for creating a pretty expressive instrument - sometimes it yielded startlingly realistic sounds, other times not so much :slight_smile: it usually sounds pretty interesting though, and definitely in the hinterland of acoustic / electronic / ??? sororities.

Sorry, I know you were asking for a modular-only example - I’m sure they are out there!


I’d love to hear an example if you have any recordings online. Sounds intriguing even though it’s far removed from what I have in mind. :blush: Thanks for taking the time to explain your process though!


not exactly what folks are talking about but the yamaha reface cp is amazing sounding for what it is (and the price). it uses samples and physical modeling at the same time and the dynamic range is incredible.

highly rec if you cant afford a real wurlitzer or rhodes or a more expensive emulation of them.


not with modular, at least not convincing strings, here is my try with osc, filter and c0 as bodyfilter, it’s more a demo of the filtering possibilities of the c0:

i’m not happy with that however, it sounds rather cheap…

the eagan matrix (synth for haken continuum) has some great string sounds.
in software kaivo, serenade ensemble for reaktor and logics sculpture are great for string sounds.


This thread needs a warning.

I was up on eBay searching for vl1 mv2s within minutes after reading it before I (luckily) caught myself and backed off. Each time it happens (and it has happened before) my resolve gets a little weaker. I’m getting worried.



Sounds like a bad violin player actually!


kilchhofer Sorry, that came out like I’m commenting on your playing, which I’m not. I’m actually very intrigued how you managed to make the playing (vibrato, glissandos) etc sound so realistic. What did you use?

I really think your recording would be more than passable in a mix. I really like how obvious the upgrade is with each element you add.


Consider the following configurations (basically different forms of feedback gain compression) in order to have more control over the interesting/chaotic/overblown region. either will work, depending on whether you have a compressor or an envelope detector.



So you’re introducing a filter and a compressor in the feedback path of the delay. Is that it? I’m not sure what the Compressor/Envelope Follower + VCA feedback is. Seems to me like you’re sidechaining the compressor with itself which (to my limited understanding) doesn’t make much sense. Care to explain?

Thank you for your time, and thanks everyone for this great thread. Once again, it has come in a much required time when I’m being very hands-on with these things.


I’ve used similar feedback paths to generate infinite non-clipping sustain in reverb/delay feedback loops. Give the feedback loop more than 1.0 gain overall, but then the envelope follower applies negative CV to the VCA to suppress it as needed.

I had been using a Bastl Dynamo with a VCA, but Maths also works. A couple nights ago I made myself an ER-301 custom unit to do the same.


haha, no problem, it sounds indeed like my daughter used to play some years ago…pretty horrible!
i used a doepfer ribbon controller for the playing.


ah i forget to add to all these patches that i now use the fantastic instruo tanh(3) as limiter inside a feedback patch: https://www.instruomodular.com/product/tanh/ extremely handy module, also sounds good as waveshaper, small and three channels, no cv though.

edit: the thing with feedback control is that you don’t want to completely limit it, the nonlinear, chaotic feedback behavior is actually needed/wanted, and can result in great “accidental” sounds. so it again comes down to control with these patches, it would be great to have the tanh with cv and then control the limiting/waveshaping with your playing (breath or bite sensor or something like that)