摘要 |
A speech communication unit (300) comprises a speech encoder (334) having a transmitter for transmitting at least a first periodic waveform and a second non-periodic waveform to a receiving speech communication unit. The speech encoder comprises a processor 480, operably coupled to the transmitter, to derive the first waveform, based only upon the previous speech portions. This improves speech quality when the signal is noisy. A Kalman filter may be used in the process, and a unit delay applied to a Kalman state to allow the previous speech portion to be used in periodic speech signal extrapolation. Optionally, if noise is not detected in the input speech signal, the processor derives the speech signal based on previous and current speech portions.
|