Presentation | 2005/7/14 Principal component analysis of spectral variations due to vocal effort in singing voice; Applications to singing synthesis Kayoko TAHARA, Toru TAKAHASHI, Masanori MORISE, Hideki BANNO, Hideki KAWAHARA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Vocal parameters (e.g. pitch, intensity, timber) are randomly fluctuating and systematically changing at the same time while singing. A series of analyses were conducted to reveal this intensity related systematic change in timbre using RWC music database and newly recorded singing sounds. The newly recorded sounds consisted of crescendo and decrescendo singing where intensities are systematically changing. They also consisted of sounds with constant intensity. A male adult with singing experience was participated in the recording. The STRAIGHT spectra of these samples were converted to third octave band levels. Then, they were converted to orthogonal components using the principal component analysis. Regression analysis of the principal components and the sound pressure levels revealed that the eigenvector of the first principal component closely represents the systematic spectral variation due to change in intensity. A preliminary test indicated that introducing spectral modifications based on this observation improves naturalness of synthesized crescendo and decrescendo singing. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | singing voice / Principal component analysis / STRAIGHT / vocal effort / voice quality |
Paper # | SP2005-38 |
Date of Issue |
Conference Information | |
Committee | SP |
---|---|
Conference Date | 2005/7/14(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Speech (SP) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Principal component analysis of spectral variations due to vocal effort in singing voice; Applications to singing synthesis |
Sub Title (in English) | |
Keyword(1) | singing voice |
Keyword(2) | Principal component analysis |
Keyword(3) | STRAIGHT |
Keyword(4) | vocal effort |
Keyword(5) | voice quality |
1st Author's Name | Kayoko TAHARA |
1st Author's Affiliation | Graduate School of Systems Engineering, Wakayama University() |
2nd Author's Name | Toru TAKAHASHI |
2nd Author's Affiliation | Graduate School of Systems Engineering, Wakayama University |
3rd Author's Name | Masanori MORISE |
3rd Author's Affiliation | Graduate School of Systems Engineering, Wakayama University |
4th Author's Name | Hideki BANNO |
4th Author's Affiliation | Department of Information Engineering, Meijo University |
5th Author's Name | Hideki KAWAHARA |
5th Author's Affiliation | Graduate School of Systems Engineering, Wakayama University |
Date | 2005/7/14 |
Paper # | SP2005-38 |
Volume (vol) | vol.105 |
Number (no) | 198 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |