Haha.
Yeah, there’s huge potential in Digital Waveguides. So underrated. It’s like Additive and/or FM/PM synthesis. Mostly unloved due to their immense flexibility and thus their inherent complexity and time consuming ‘programming’. But even relatively simple implementations can be very usefull.
The subject of Digital Waveguides (and much more on sound and DSP) has been deeply researched by Julius O. Smith III (et al) at CCRMA at Stanford University. Much of this research is published in great detail, free of charge. Some of it may be licenced or even patented.
The VCV Rack Development guide on DSP at
also points to his homepage and work at CCMRA Stanford
Julius Orion Smith III
Professor Emeritus
Music and by courtesy Electrical Engineering
CCRMA Home Page
JOS Home Page
Global JOS Index
JOS-Hosted On-Line Publications
Taxonomy of Digital Synthesis Techniques
And CCRMA Stanford and Julius O. Smith are not the only ones researching and publishing great stuff…
Anyway…
Maybe just start with the basics of a small (1 x 1, 2 x 2…max 8 x 8) 2D Waveguide Mesh and its variables and configurations. Waveguides are computionally very efficient (in terms of Physical Modelling) but nonetheless, the computational resources needed can soon multiply (exponentially) with size and complexity of the mesh.
Physical Modeling with the 2-D Digital Waveguide Mesh
Recently I (re)tried my luck at plugin development. I actually did set up a VCV Development environment on my machine and messed with some code. The main trigger to do this was because as of last November (2022) AI (Large Language Models, notably ChatGPT at that time) seemed to be able to fill in the gaps in my knowledge and c++ coding skills…
I have been messing with the internals of some existing Modal/Waveguide plugins (e.g. NOI Sinesis, Sckitam WaveguideDelay, docB MVerb, Frequency Domain Dance This Mesh Around and such). But I just lack the technical skills (and time to invest) to realize my ideas (GUI and internals). Currently I can’t even adapt and combine existing modules/code with the help of AI LLM’s and resources on the Wild Wild Web towards something fundamentally adding something to what already exists…
Altough, about AI/LLM’s and developing/coding
The current AI’s/LLM’s are a lot more plentiful and powerful then a year or even months ago and are still making progress at exponential (and alarming) rates. They are not only becoming evermore helpful as a knowledge base and code generator (and verifier/optimizer). They are also more and more able to (cross) translate from/to other languages (both human spoken and programming languages). Thus opening up knowledge and projects in other languages then your own and c++.
Maybe I will give it another try later…by just using AI, Natural Language Processing (NLP)/LLM prompts and other modes. And just explain what i want and sketching GUI’s and have an AI generate the code. And use Voice Recognition/Processing modes to complain in realtime about unexpected results…