Artificial Intelligence (AI) is transforming the world and how we make music. This doctoral research uses cutting-edge computer techniques in machine learning to create digital emulations of acoustic instruments and analog effects. These emulations are designed to capture these real-world devices’ unique and complex sounds.
Why is this important? Acoustic instruments and electronic hardware may be inaccessible because they are costly and bulky. However, digital versions often lack the rich and vibrant sound of their physical counterparts. Think about the difference between listening to a live piano performance and a digital keyboard trying to imitate it.
This research aims to push the current boundaries and make digital music sound more realistic. State-of-the-art AI technologies are not only enhancing the realism of digital emulations of musical devices but also creating new audio effects, opening novel creative opportunities.
The focus is on creating efficient, responsive, and interactive emulations, enabling musicians to perform in real-time without delays or compromised sound quality. The techniques are also designed to work across various instruments and effects, making them versatile tools for emulating musical devices. Imagine a future where digital production tools are just as vibrant and nuanced as their real counterparts. This research is a step towards that reality, offering exciting possibilities for musicians, composers, producers, and music enthusiasts
Riccardo Simionato succesfully defended his thesis on 28 February 2025.
Trial lecture
The designated topic: "Audio Engineering meets AI: Dataset and Model Lifecycles in Real-Time, Small Models"
Evaluation committee
- Associate Professor Federico Fontana, University of Udine (first opponent)
- Associate Professor Cumhur Erkut , Aalborg University (second opponent)
- Associate Professor Tejaswinee Kelkar, University of Oslo (committee administrator)
Chair of the defence
- Associate Professor Hans T. Zeiner-Henriksen
Supervisors
- Associate Professor Stefano Fasciani, University of Oslo
- Professor Sverre Holm, University of Oslo (co-supervisor)