Development of rhythm practice supporting system with real-time onset detection

Abstract

We are developing a rhythm practice support system with onset detection. The system produces guiding metronome sounds and processes the performed sound to detect onsets in real time. Spectral flux and dynamic thresholding are used to find onsets in the audio signal. We made three types of prototype display for the system; a survey of user display preference was conducted. An evaluation of the accuracy of the onset detection is also provided.