Modeling musical devices—whether instruments or effects, acoustic or electronic—represents a dynamic research area with significant scientific relevance, rich historical foundations, and artistic implications. This multidisciplinary field bridges musical acoustics, digital signal processing, and increasingly, machine learning approaches. The seminar examines state-of-the-art neural network applications in modeling musical devices, from analog effects to acoustic instruments, emphasizing low-latency and interactive emulations for real-time performance. Through case studies of compressor effects and acoustic piano modeling, the presentation demonstrates advances in neural architectures for effect and instrument modeling.
- Simionato, R., & Fasciani, S. (2025). Modeling time-variant responses of optical compressors with selective state space models. Journal of Audio Engineering Society, 73(3), 144–165.
- Simionato, R., & Fasciani, S. (2025). Sines, transient, noise neural modeling of piano notes. Frontiers in Signal Processing, 4, 1494864.
Presenter : Riccardo Simionato, postdoc at SCRIME @ LaBRI, https://riccardovib.github.io/