Vocal Expression Transfer for Singing Voice Synthesizer

Back to members

We demonstrate singing style transfer for Vocaloid.
First, vocal expressions of a singer are transcribed using our method. The collection of the expression parameters is the singer's library. Secondly, the vocal expressions are attached to Vocaloid (VSQ) file.

Process of Vocal Expression Transcription

Transfering Vocal Expressions

Kobushi, glissdown, and glissup are transfered as PIT parameter, and vibrato is transfered as Vibrato parameter of Vocaloid.
"Spitz" is a Japanese pop group and "Misora Hibari" is a Japanese enka singer. Now, Enka is a traditional Japanese music genre, which has kobushi and deep vibrato.

Rules for Transfering Vocal Expressions

We designed a simple rule for transfering vocal expressions.

Sound

Music : RWC-MDB-P-2001 No. 7 "PROLOGUE" [1]

Synthesizer	VOCALOID2 (Megpoid)	Cevio (Sato Sasara)
No expressions
"Spitz"
"Misora Hibari"

Reference

1. Masataka Goto, Hiroki Hashiguchi, Takuichi Nishimura, and Ryuichi Oka: RWC Music Database: Popular, Classical, and Jazz Music Databases, Proceedings of the 3rd International Conference on Music Information Retrieval (ISMIR 2002), pp.287-288, October 2002.