Authors:
Hidefumi Ohmura
1
;
Takuro Shibayama
2
;
Keiji Hirata
3
and
Satoshi Tojo
4
Affiliations:
1
Department of Information Sciences, Tokyo University of Science, 2641 Yamazaki, Noda-shi, Chiba, Japan
;
2
Department of Information Systems and Design, Tokyo Denki University, Ishizaka, Hatoyama-cho, Hikigun, Saitama, Japan
;
3
Department of Complex and Intelligent Systems, Future University Hakodate, 116-2, Kamedanakano-cho, Hakodate-shi, Hokkaido, Japan
;
4
Graduate School of Information Science, Japan Advanced Institute of Science and Technology,
Keyword(s):
Music, Melody, Lattice Space, GMM, EM Algorithm.
Abstract:
Music is organized by simple physical structures, such as the relationship between the frequencies of tones. We have focused on the frequency ratio between notes and have proposed lattice spaces, which express the ratios of pitches and pulses. Agents produce melodies using distributions in the lattice spaces. In this study, we upgrade the system to analyze existing music. Therefore, the system can obtain the distribution of the pitch in the pitch lattice space and create melodies. We confirm that the system fits the musical features, such as modes and scales of the existing music as GMM. The probability density function in the pitch lattice space is suggested to be suitable for expressing the primitive musical structure of the pitch. However, there are a few challenges of not adapting a 12-equal temperament and dynamic variation of the mode; in this study, we focus on these challenges.