To provide a method and device for encoding an acoustic signal, which is capable of realizing faithful reproduction of a consonant component by using a sound source reproduced with a limited number of frequencies by utilizing human auditory characteristics.
A digital acoustic signal constituted of time series sample strings is analyzed per unit section to obtain spectral intensity for a plurality of frequencies, and then the frequencies are divided to a plurality of groups by bands, and a frequency having a maximum spectral intensity is searched for within a range from a lower limit pitch to an upper limit pitch with respect to each band (S22), and the spectral intensity for frequencies other than the frequency having the maximum spectral intensity is attenuated (S23). With respect to each band, a result from attenuation of the spectral intensity other than the maximum spectral intensity is used to perform encoding.
| JP2008516288A | 2008-05-15 | |||
| JP2011090189A | 2011-05-06 | |||
| JP2001007704A | 2001-01-12 |
JPN6014049768; 茂出木 敏雄 Ｔｏｓｈｉｏ Ｍｏｄｅｇｉ: 'オーディオ-ＭＩＤＩ符号化技術「オート符」の改良と楽器音によるテキスト音声合成システムへの応用' 第７２回（平成２２年）全国大会講演論文集（２） 人工知能と認知科学 , 20100308, ｐ．２-７７５〜２-７７６, 社団法人情報処理学会
CSNG201000619006; 茂出木 敏雄 Ｔｏｓｈｉｏ Ｍｏｄｅｇｉ: 'ＭＩＤＩ符号化ツール「オート符」を用いた音素ＭＩＤＩコードの設計と楽器音による音声合成機能の実現' 電気学会論文誌Ｃ Ｖｏｌ．１３０ Ｎｏ．７ ＩＥＥＪ , 20100701, ｐ．１１５９-１１６７, （社）電気学会 Ｔｈｅ Ｉｎｓｔｉｔｕｔｅ ｏｆ Ｅｌｅｃｔｒｉｃａｌ Ｅｎｇｉｎｅ
CSNJ201010051376; 茂出木 敏雄 Ｔｏｓｈｉｏ Ｍｏｄｅｇｉ: 'オーディオ-ＭＩＤＩ符号化技術「オート符」の改良と楽器音によるテキスト音声合成システムへの応用' 第７２回（平成２２年）全国大会講演論文集（２） 人工知能と認知科学 , 20100308, ｐ．２-７７５〜２-７７６, 社団法人情報処理学会
伊藤 裕介
後藤 直樹
深町 圭子
藤枡 裕実
金山 聡
Next Patent: CAVITY RESONATOR