| 摘要 |
PURPOSE:To eliminate monotonousness and a mechanical sound, and synthesize a natural speech by giving specific fluctuations to a synthesized sound. CONSTITUTION:This speech synthesizing device has a speech synthesizing means 3 which outputs the contents of an input character text 1 as a synthesized speech waveform 2. Then the device is provided with a 1/f fluctuation generation part 4 which generates random fluctuations having such frequency characteristics that the power in inverse proportion to the frequency, and the fluctuations are added to an output speech. Further, the fluctuations can be added to the pitch of the output speech, the amplitude of the output speech, or the spectrum of the output speech. Consequently, the speech is neither monotonous nor mechanical and becomes natural. |
