| 摘要 |
PURPOSE: To effectively code the speech of a low data speed by changing the interval between channels corresponding to the pitch of the voice of a speaker. CONSTITUTION: The speech is sampled, a series of discrete samples are obtained and a series of frames over the respective plural samples are obtained from the samples. Respective sample frames are analyzed, a set of frequency components provided with individual amplitudes and phases are extracted and the components are tracked from one frame to the next frame. The value of the components from one frame to the next frame is interpolated, the parameter display of a waveform is obtained and a set of sine waves for which a synthetic speech waveform corresponds to the interpolation value of the parameter display is generated. For instance, a set of the linearly separated frequencies of a fundamental band and the other set of the logarithmically separated frequencies of the frequency area of a higher order are used for coding the amplitude in a transmitter, and in a receiver, interpolation is linearly performed between channel amplitudes and another amplitude envelope is constituted. It is sampled by pitch harmonic waves and a set of sine wave amplitudes used for synthesis is generated. |
