Document Type
Article
Publication Date
8-28-2025
Abstract
Accurate measurement of note-to-note transitions is essential for analyzing articulation in clarinet performance. Traditional methods rely on either subjective amplitude thresholds—such as the time between 5% and 95% RMS levels—or direct measurement of tongue-reed contact time using reed-mounted sensors. These approaches are limited by their dependence on user-defined parameters or invasive hardware. This study proposes a computational alternative: ΔT, a curvature-based metric defined as the time interval between surrounding minima in the second derivative of the mouthpiece pressure envelope. Using data from a sensor-equipped mouthpiece (SEM), we compare ΔT to both threshold-based timing (Tt) and tongue contact duration (Tc) across portato and staccato articulations. Our findings show that ΔT closely tracks both Tt and Tc in structured articulations, with minimal absolute difference and robust repeatability. These results support the use of ΔT as a non-invasive, objective, and reliable estimate of transition duration, enabling broader application in performance analysis, pedagogy, and real-time feedback systems. This research was funded by: National Science Foundation Grant 2109932.
Published In
Griffin, Max, Coyle, Whitney L. and Makayle S. Kellison. “Computational Detection of Articulation Transients in Clarinet Performance.” Proceedings of Meetings on Acoustics 58, no. 1 (2025): 1-9. https://doi.org/10.1121/2.0002074.
Publication Title
Proceedings of Meetings on Acoustics
DOI
https://doi.org/10.1121/2.0002074
Comments
Originally published in the Proceedings of Meetings on Acoustics.