| 摘要 |
PURPOSE:To eliminate the need for a long time for generation and to standardize the quality by extracting only a voiced part from a digital signal and adding previously set voiceless parts before and after the voiced part unconditionally. CONSTITUTION:A speech signal of a word or syllable is inputted continuously into an analog-digital conversion part 2 from a sound source 1 such as a tape recorder, etc., while containing voiced and voiceless parts, and then converted into the digital signal, which is sent to a voiced part detection part 3. The voiced part detection part 3 extracts only voiced parts from the speech signal and the respective extracted voiced parts correspond to the word or syllable and are sent to a voiceless part addition part 4, which adds the preset voiceless parts before and after the extracted voiced parts unconditionally. Thus, a voiceless part A is added before the voiced parts and a voiceless part B is added after the voiced parts. Consequently, the word and syllable which are used for a sound recording editing type speech response device are written on a storage device efficiently with good quality. |
