Guitar synthesis improved with editable note and string control
Expressive Guitar Synthesis via Controllable Latent Flow Matching
Sound
Summary
Creating realistic guitar sounds with computers is hard because simply knowing what notes to play doesn’t say how to play them expressively. The authors propose a two-part system where the first part predicts how each guitar string's pitch and loudness should change over time based on the note instructions, and the second part turns these predictions into actual guitar sounds. This approach allows users to edit the details of the guitar performance, like string articulations, which was difficult before. Their experiments show that this method produces better-quality guitar sounds and offers more control than previous single-step methods.
What this means in practice
- •For music software developers: Add precise, editable guitar note and string articulation controls to virtual instrument software.
- •For audio plugin designers: Create guitar synthesis plugins that allow musicians to customize performance expressiveness per string and note.$Commercial implications: Enables development of advanced guitar synthesizer plugins with fine-grained expressive controls attractive to musicians.
Authors
Hsueh-Wei Fu, Yen-Tung Yeh, Yi-Hsuan Yang
Abstract
Realistic guitar synthesis requires both audio fidelity and interpretable control over how notes are performed. The input timed MIDI events decide what is played but not how, leaving the continuous within-note decisions of guitar performance--such as pitch inflection and per-string articulation--unspecified. A model that directly maps MIDI to audio must therefore realize the performance implicitly, and users are left with limited control over the synthesized performance. We therefore propose a two-stage guitar synthesis pipeline that predicts editable, per-string $f_0$ and amplitude trajectories from timed note events and note-level dynamics, and renders them into audio using latent flow matching. Experimental results show that the proposed system outperforms a matched single-stage counterpart and DDSP-Guitar in control fidelity and perceived quality, while keeping the controls editable. Sound examples can be found online.