Realtime conversion of growl-type voice qualities based on modulation and approximate time-varying filtering driven by a non-linear oscillator: Formulation

スポンサーリンク

概要

• 論文の詳細を見る

• A formulation of voice conversion to add growl-like voice qualities to singing voices is proposed based on our findings of features in such singing performances. The proposed method does not consist of any analysis and synthesis stage(s). A preliminary implementation using Matlab demonstrated that its throughput is faster than realtime. The proposed formulation provides not only post processing capabilities of rendering styles of existing performances to recorded materials but also realtime capabilities of adding growl-like voice qualities in live performances.

• 2014-02-16

著者

• Hideki Kawahara

  Wakayama University

• Shohei Mizobuchi

  Wakayama University

• Masanori Morise

  University of Yamanashi

• Ryuichi Nisimura

  Wakayama University

• Toshio Irino

  Wakayama University

関連論文

• Realtime conversion of growl-type voice qualities based on modulation and approximate time-varying filtering driven by a non-linear oscillator: Formulation

スポンサーリンク