가야금 농현의 음향학적 구조와 디지털 음고 검출 알고리즘의 한계 = Acoustical Structure of Gayageum Nonghyeon and Limitations of Digital Pitch Detection Algorithm

다국어 초록 (Multilingual Abstract)

This study analyzes the acoustical characteristics of Gayageum Nonghyeon and examines the interpretive limitations encountered when applying existing digital pitch detection algorithms to this technique. Unlike Western vibrato, which involves symmetrical fluctuations around a central pitch, Nonghyeon is performed by increasing string tension, resulting in an asymmetric pitch trajectory that fluctuates only upward from the floor pitch.
Experimental analysis of Gayageum and cello performances using Praat and Melodyne revealed that pitch estimation methods based on mean or center values tend to recognize Nonghyeon as being upwardly biased compared to the performer's intended pitch. This occurs because conventional algorithms fail to account for the fact that the perceptually recognized reference pitch in Korean traditional music (Gugak) is anchored at the floor (base note) rather than the average.
To address these limitations, this study proposes the floor pitch, lower envelope, and statistical mode as alternative criteria for Gayageum pitch analysis. We discuss how these metrics more accurately reflect the structural characteristics of Nonghyeon and the performer’s intent. These findings provide a foundational framework for improving Gugak pitch analysis and developing future traditional music-based acoustic technologies and AI applications.