Physical modeling with existing chains\units?

hello there! the wait is almost over! i’m really excited. i’ve been disciplined enough to not go crazy in anticipation in those last two months but now excitement is taking over :slight_smile:
i have a question:
i saw in the todo list that there’s a waveguide unit in the roadmap but i was wondering if there’s a way at this point in time to build some kind of modal synthesis\physical modeling custom unit with the chains and units we got. i mean something beyond a simple karplus strong patch.
i was thinking maybe a bank of resonant filters? will i be able to just ping them with a short click, pulse or envelope-shaped noise? i believe building a feedback loop is very straight forward just with mixer units. has anyone already tried something similar?
then i think i’d place some control on the harmonic ratio\placement of bands and have some control on their individual decays (maybe in a global way, with different ranges on each band etc…) and maybe other modulations to bend the harmonic ratios etc…
i was thinking something similar to reaktor prism (which i find pretty good!)
what do you think?

4 Likes

Yeah, I like Prism too.

@NeilParfitt’s Karplus-Strong video is the closest I’ve seen to physical modeling synthesis on the ER-301.

I tried for quite a while and was never able to get the 1 v/o tuning dialed in on it either.

As far as resonant banks, there’s no band pass filter yet. You could of course build one up out of a low and high pass filter in series. But for that kind of thing I’d think you’d want control over the BW whereas the ladder HPF and LPF are fixed slope.

So I think it might be tough at this point. But the firmware is still young. :slight_smile:

3 Likes

thanks Joe!!! well, i’ll try with multiple k-s lines and some extended control :slight_smile:

1 Like

I certainly wouldn’t discourage you from trying! Will be interested to see what you come up with (this or otherwise).

With the right envelopes Evil Twin can sound very ‘physical’ especially if you stack another sound source in there through the ext input :slight_smile:

1 Like

thanks Joe and Kel! about 10 days\two weeks left for my 301 to land here…my mind is sort of trying to become a er-301 emulator :smiley:

3 Likes

Haha. I know exactly what you mean. Did the same thing myself. :slight_smile:

2 Likes

I still do this now, even though I have the physical units, I am not always in front of them :smile:

1 Like

good physical modelling isn’t possible with the er301 alone, and synthesizers like prism, kaivo or chromaphone are such complex synths, i don’t think it’s possible to build something like that within modular, especially not polyphonic. the heart of prism is a modal bank, consisting of many many resonant filters which play together nicely to build a sort of body of a sound which you excite. even if you have dozens of resonant bandpass filters you still would need an “intelligent” interface to modulate them all simultaneously.
crude karplus strong or resonant filters however are certainly possible to build, however it’s difficult and not so practical since good physical modelling plays with nonlinearities and feedback paths to achieve a natural sound. i remember however that brian has added a waveguide model onto the “eventually” tab in the o/d wiki page, that definitely would be awesome!!!

yes i’m aware of the complexity of good modal synths, but i was thinking something rough. and i think that organizing controllers for multiple bands with clever scaling is something that the 301 is really capable of! maybe not 64 partials :smiley: but something crude like 12 partials with a decay envelope for each band with a global control and another control (vca?) for proportionally shortening the decay the higher the harmonics…something like that :slight_smile: oh, and a filtered feedback path too!
(p.s.: by chance i was just listening to your “dersu”, really awesome record!!!)

1 Like

waiting for my 301 to ship i’m prototyping a crude modal synth on reaktor avoiding the modal bank and doing it with single resonant bandpass filters, noise generator and impulse generator through envelopes and feedback. i’m now playing with just 8 partials and i must say i like the results! obviously it’s very far from the organic and complex sound of something like prism, sculpture or even collision but it’s something i find musically useful! so i’m faithful that i’ll be able to do something similar in the 301. in some ways the parameter control and the general UI of the 301 seems to me even easier to manage than some aspects of reaktor. the part i’m having the more difficulties is the feedback path and keeping the gain consistent. but i’m very, very happy with the results with just 8 partials. the huge difference will be polyphony of course. on reaktor (with some audio voice combiners) i can easily switch the entire beast to polyphonic mode, in the 301 i’ll be more restrained to monophonic sound but…

3 Likes

sounds good! with 8 bandpass filters you can do alot, i work often with “just” the four of a res-4. but as joe already mentioned above there’s no bandpass filter for the er301 yet and i find it quite hard to work with the existing lp and hp combination, the resonance behavior isn’t suited for resonator-type sounds…eg the sweet spot of decaying resonance and self-oscillation is veeery small.

1 Like

ahhh yess! i was missing that point…well i’m quite confident a bp filter will show up sooner or later :slight_smile:
i found a way to cheat a bit on the feedback side…just put a * (vca) after the feedback (before a crossfader that blends exciter signal with feedback) with a simple envelope controlling it…not optimal but way more controlled…

I’m resurrecting this thread with the hopes that v0.4.x makes for easier physical modelling. Additionally we have @Joe’s bespoke BPF now!

My intentions are to investigate the following:

  1. Mapping of v/o for tuned delay line
  2. Above, but using middle layer
  3. Use of external feedback, especially resonant EQ / bandpass filters
  4. Use of internal feedback
4 Likes

My interest would be in sub-audio-rate physical modeling, that is e.g.

  • Newtons Mechanics (see ADDAC Marble Physics, Bouncing Balls et al.)
  • Liquid Dynamics (see pmpd )
    Inside PMPD are many different “physics” - a port would be my dream.
    Would be amazing to hear this signals control parameters inside ER301 or going into an external quantizer etc. :slight_smile:
    btw the Hordjik CU-ports are going already in this direction, even though from another angle.
1 Like

thanks for checking the hordijk ports. my research is going in the direction of trying to build with the most low level approach without resorting to the middle layer. sort of learning analog computing in a digital environment
check also the Math Perc cu for another example of that approach.
here the attempt of a benjolin port

1 Like