Study in npj Acoustics presents the first physics-based computational violin simulation, modeling a real Stradivarius via CT scan and finite element analysis to produce realistic pizzicato sound.
Key Takeaways
Built from CT scans of a 1715 Stradivarius, the model divides the violin and surrounding air into millions of finite elements with per-material physics equations.
Produces sound from first principles, not sample libraries, enabling luthiers to hear the effect of changing wood type or plate thickness before building anything.
Currently limited to pizzicato (plucked strings); bowed string simulation is explicitly flagged as a harder, unsolved problem due to nonlinear bow-string interaction.
Demonstrated on Bach’s Fugue in G Minor and Daisy Bell; uniform pluck timing makes individual notes sound mechanical compared to human performance.
Intended as a design tool to accelerate iteration: current luthier workflow requires building a full instrument before evaluating sound changes.
Hacker News Comment Review
Commenters noted the physics modeling space is not new: commercial tools like Audio Modeling have offered physical modeling synthesis for years, with plugins under 15MB versus tens of GB of sample libraries.
The bowing omission drew skepticism, since bow-string slip is the defining nonlinear behavior of violins and the primary mode luthiers evaluate; pizzicato is a secondary use case.
Physical modeling plugins require heavy real-time parameter control via DAW automation or MIDI controllers to sound realistic, making them harder to use than samplers despite size advantages.
Notable Comments
@shooly: flags Audio Modeling as prior art and notes the disk-size vs. usability tradeoff of physical modeling over sampling.
