ABSTRACT

This study aimed to establish reference criteria for learners and performers by defining nonghyun—a core expressive element in Korean traditional music—within rhythmic sobak units. Using deep learning-based analysis with pitch extraction (CREPE) and Fast Fourier Transform (FFT), gayageum sanjo performances were analyzed and visualized. Within clearly periodic rhythmic units, nonghyun segments were identified and measured in pitch, depth, speed, and frequency. The analysis showed that Jinyangjo exhibited about three nonghyun per beat and Jungmori about two, maintaining rhythmic periodicity despite tonal or phrasing variations. Future research should refine segmentation criteria and extend to diverse sanjo repertoires and performers.

Keyword :