From Textural Counterpoint to Feature Encoding: A Multi-Dimensional Machine Representation Study of Haydn's "The Lark" Integrating Electroacoustic Analysis

This paper studies the first movement of Haydn’s String Quartet in D Major, “The Lark” (Op. 64, No. 5), as a model for machine-readable musical interaction. It combines qualitative auditory analysis with DAW-based spectrum and dynamics measurements, then proposes event-based timestamps instead of a conventional quantization grid. Acoustic features are organized into a “Role-Aware Encoding” intended to represent leading voices, accompaniment, micro-timing, and interactive texture. The paper frames this representation as a possible foundation for human-computer collaborative composition and role-sensitive symbolic music generation. The abstract presents a methodology and theoretical argument, but does not report detailed benchmarks or generation experiments.