Expressive coupling and textural dissociation in classical piano performance: evidence from the ASAP dataset

Area of Science:

• Music Performance Science
• Computational Musicology
• Cognitive Science of Music

Background:

• Empirical methods are needed to link musical interpretation with measurable behavior in skilled performance.
• Analyzing expressive timing and dynamics in classical piano performance requires sophisticated data analysis.

Purpose of the Study:

• To investigate the expressive coupling between timing flexibility and dynamic shaping in piano performance.
• To examine the textural dissociation of timing behavior from surface note density.
• To compare expressive behaviors across different musical repertoires.

Main Methods:

• Utilized the Aligned Scores and Performances dataset for classical piano performances.
• Extracted features: beat interval variability, local tempo change, MIDI velocity dynamic range, and note density.
• Grouped composer labels into broad repertoire categories (e.g., Baroque, Romantic) for comparative analysis.

Main Results:

• Identified significant differences in timing and dynamics across musical repertoires, with Romantic music exhibiting more flexible timing and wider dynamic range.
• Confirmed a strong expressive coupling between timing variability and dynamic range.
• Demonstrated that global timing flexibility is dissociated from surface note density.

Conclusions:

• Expressive coupling (timing and dynamics) and textural dissociation (timing from note density) are measurable features of pianistic interpretation.
• Temporal and dynamic control in piano performance operate together but are distinct from surface note density.
• Feature-based prediction models can reliably capture these performance nuances.