KVR Audio

Xenos · Post by **Xenos** » Wed Mar 12, 2014 11:01 pm

I've thought for a while that pure physical modeling synthesis is capable of so much more than just emulating real instruments. Wasn't the Hartmann Neuron hardware synth supposed to be an expansion on that concept?

Wouldn't it be cool if a dev took the core physical modeling concept and abandoned the idea of emulation altogether, coupled with an efficient, streamlined workflow? Sure, Tassman is quite capable, but it's a bear to use (just my own opinion).

I really like inherent cold, rough sound you get from physical modeling when you deviate from trying to make the same old guitars and violins. It's unique. Definitely distinctive from VA, FM, granular or additive. IMHO, to really use physical modeling as a form of abstract synthesis would require access to more advanced and specific parameters than what is typically offered in a PM synth.

Take a selection of different exciters - hammer, membrane, plucked string, bowed string, brass mouthpiece, reed mouthpiece, flute mouthpiece, and the typical synth waveforms + noise. Run them through a wide selection of resonators, which are then run through some analog style filters, plus effects at the end. Here's where it gets cool: How about modulating the SIZE of the chosen resonator and the MATERIAL model with LFOs, envelopes, modwheel, etc.? How about being able to modulate some of the exciter parameters? Maybe have 2 layers of exciter --> resonator --> output instead of one, plus a 2nd filter for much more sonic possibilities. You could have one layer with a typical saw waveform giving a steady tone (run through modulating resonators for interesting flange-like effects), and another layer an octave higher using a bowed waveform for a dirty, scraping pad sound. Have each layer run through its own filter. Add a delay at the end, set to a high feedback setting, and have the delay speed modulated by an LFO at a low setting. Should be able to get some interesting, metallic, moving soundscapes that way.

electro · Post by **electro** » Wed Mar 12, 2014 11:16 pm

Why do you think PM has still yet to produce Acoustic instruments as convincing as samples? Moddart Pianoteq comes the closest.

Xenos · Post by **Xenos** » Wed Mar 12, 2014 11:28 pm

electro wrote:Why do you think PM has still yet to produce Acoustic instruments as convincing as samples? Moddart Pianoteq comes the closest.

Oh I have no doubts about that. Sorry if I gave the wrong impression. PM is definitely good at what it advertises itself as being. I just happen to really like the "dark side" of PM synthesis and see potential in that. The "wrong" settings on a PM synth can sound harsh and grating, or just plain weird, and I like it. There's an abstract, non-acoustic side of PM synthesis and I thought it'd be really cool to expand on that.

Sendy · Post by **Sendy** » Wed Mar 12, 2014 11:34 pm

Kaivo is somewhat in that ballpark. Though personally I find it more useful for creating sound effects, ambiences and chance-based sample loops, rather than using it with midi for riffs and stuff - true Buchla stylee

EnGee · Post by **EnGee** » Thu Mar 13, 2014 12:04 am

I really like a lot the dreamy pads in NI Prism. I'm interested as well in the wide range of weird sounds it produces, from metallic hits to those cold dark pads (I'm glad that I kept it!).

I actually like its sound more than Logic Sculpture but I just need enough time to spend learning it.

electro · Post by **electro** » Thu Mar 13, 2014 12:08 am

The whole point of PM I thought was to capture the timbre of real acoustic instruments and detailed response to continuous controllers that sample based instruments could not provide. AAS Tassman is good for the experimental side of PM.

glokraw · Post by **glokraw** » Thu Mar 13, 2014 9:44 pm

xoxos has modeled/synthesized many aspects of sound, so many that
the sum might be considered far beyond emulation,
even if a precise targeted emulation,
was achieved, or was the initial goal.
Cheers

Burillo · Post by **Burillo** » Thu Mar 13, 2014 9:47 pm

there was an instrument which i think was called sylfid or something... it is a "physically modeled" instrument that doesn't actually exist in the real world (i.e. entirely original concept based on physical modeling).

glokraw · Post by **glokraw** » Thu Mar 13, 2014 10:54 pm

What is unreal about the world a 'sylfid' is created in?
Everything started/starts out out as an original concept.
Lots of designs are not modeled after a particular item,
but then become one.

http://www.prymaxevintage.com/bmf-effec ... istortion/

Cheers

aMUSEd · Post by **aMUSEd** » Thu Mar 13, 2014 11:23 pm

electro wrote:The whole point of PM I thought was to capture the timbre of real acoustic instruments and detailed response to continuous controllers that sample based instruments could not provide.

That's never how I've seen it, I've always seen emulation as just the starting point not the end in itself. Once you are able to fully model the responsiveness of a real instrument it means you then have total control over timbre and expression, and once you have that you can go way beyond what the original instrument offers to do things like morph between instrument types or create unique instruments of your own. Using Modelonia and Kore for example I created many patches that allow me to morph in realtime between say a flute and guitar sound, or sax and sitar and explore some amazing sound spaces in between.

BasariStudios · Post by **BasariStudios** » Fri Mar 14, 2014 4:53 am

PM was/is one of my specialties, it actually surpassed it self in acoustic stuff. Try Yamaha VL, Korg Prophecy, Z1, Korg MOSS Board on the Tritons. I don't love nothing more then PM, worked with PM for years and still do but i still have not found a software that can actually do real Acoustic Physical Modeling such as Yamaha's VL and the Korg stuff. There is no PM VST Synth for me yet. Maybe Sculpture? Prism? But still, not even close yet.

realmarco · Post by **realmarco** » Fri Mar 14, 2014 4:59 am

https://www.applied-acoustics.com/tassman/overview/

BasariStudios · Post by **BasariStudios** » Fri Mar 14, 2014 5:11 am

Nah, Tassman is not even close to anything to pysical modeling.

Uncle E · Post by **Uncle E** » Fri Mar 14, 2014 6:43 am

glokraw wrote:xoxos has modeled/synthesized many aspects of sound, so many that
the sum might be considered far beyond emulation,
even if a precise targeted emulation,
was achieved, or was the initial goal.
Cheers

+1. XOXO's did it.

I quite liked the Korg Z1 for being such a poor physical modeling synth, I was able to coax sounds out of it that I couldn't get out of better physical modelers.

whyterabbyt · Post by **whyterabbyt** » Fri Mar 14, 2014 10:05 am

BasariStudios wrote:Nah, Tassman is not even close to anything to pysical modeling.

Ridiculous statement; physical modelling is a technique, not a result. And Tassman certainly includes physical modelling components, acoustic and electronic.

The conflation of 'physical modelling' with 'stuff that sounds like acoustic instruments' is missing the point of what PM actually is; a catch-all term for a fairly wide variety of synthesis algorithms which have their basis in the simulation of real-world phenomena or behaviours. That includes models of electronic circuit behaviour, such as filters and oscillators.

Just because the 'parts' of a synth involve 'physical modelling' doesnt mean that the larger synth is a model of some specific thing, though, let alone some specific acoustic thing. An example would be Rapture, which uses waveguides, a long-established PM method.

Physical Modeling synths. More than just emulation?