著者

中山 一郎 岡田 稔枝 中川 みかほ

出版者

一般社団法人情報処理学会

雑誌

情報処理学会研究報告音楽情報科学(MUS) (ISSN:09196072)

巻号頁・発行日

vol.1999, no.16, pp.5-8, 1999-02-18

---

これまで筆者らは、発声時に発声者自身が聴取する音声(自己聴取音)の音色を、自己聴取音に相当する音声を発声者がイコライザとアッテネータを操作して、自己聴取音と出来るだけ似かよった音声(シミュレート音)を作成する方法(遅延音帰還法)を用いて、周波数軸上で定量化してきた。しかしながら、自己聴取音は発声者本人しか聴取できないが故に、シミュレート音との類似性を客観的には評価できなかった。そこで本研究では、男性の話し声の母音/a/を対象に、1)自己聴取音とシミュレート音、及び、2)シミュレート音とその音色を実験者が意図的に変化させて作成した音声(加工音)、との2種類の類似度評価実験を行うことによって、自己聴取音とシミュレート音の類似性が原理的には客観的に評価できること、及びその類似性が十分に高いことを明らかにした。What sort of voice timbre do we perceive in autophonic production, hearing our own voice during phonation, and can we measure the timbre itself quantitatively? The authors have showed that the timbre can be examined quantitatively as a function of frequency using a delayed feedback method created by the author, in which a sound was generated by the subjects themselves so as to simulate the sound through a loud-speaker as accurately as possible in its timbre and loudness to the sound perceived during phonation. However, the judgement of similarity of voice timbre between the perceived and the simulated sounds was left to the subjects themselves, not objectively, because the voice timbre in autophonic production cannot be examined by anyone other than the speaker oneself. The purpose of the present study is to show a method to evaluate the similarity between the simulated and perceived sounds objectively. When male subjects phonated a vowel/a/ in the spoken mode. Two similarity judgement tests were conducted, one (TestA) between the perceived and the simulated ones, and the other (TestB) between the simulated and the processing sounds, with an amplitude of a frequency band of the simulated sound altered, using the same criterion of the judgement of the similarity as in TestA. Through these tests, it was found objectively that the similarity between the simulated and perceived sounds was fairly well.