Fast Differentiable Modal Simulation of Non-linear Strings, Membranes, and Plates
This work addresses a bottleneck in acoustics and audio synthesis by improving simulation speed and enabling inverse modelling, though it is incremental as it builds on existing modal methods with new computational optimizations.
The paper tackles the computational inefficiency and lack of differentiability in modal simulations of non-linear strings, membranes, and plates, introducing a fast, GPU-accelerated framework that significantly outperforms existing implementations and enables gradient-based inverse modelling to recover physical parameters from data.
Modal methods for simulating vibrations of strings, membranes, and plates are widely used in acoustics and physically informed audio synthesis. However, traditional implementations, particularly for non-linear models like the von Kármán plate, are computationally demanding and lack differentiability, limiting inverse modelling and real-time applications. We introduce a fast, differentiable, GPU-accelerated modal framework built with the JAX library, providing efficient simulations and enabling gradient-based inverse modelling. Benchmarks show that our approach significantly outperforms CPU and GPU-based implementations, particularly for simulations with many modes. Inverse modelling experiments demonstrate that our approach can recover physical parameters, including tension, stiffness, and geometry, from both synthetic and experimental data. Although fitting physical parameters is more sensitive to initialisation compared to other methods, it provides greater interpretability and more compact parameterisation. The code is released as open source to support future research and applications in differentiable physical modelling and sound synthesis.