| 摘要 |
A code-excited linear-predictive (CELP) coder for speech or audio transmission at compressed (e.g., 16 kb/s) data rates is adapted for low-delay (e.g., less than five ms. per vector) coding by performing spectral analysis of at least a portion of a previous frame of simulated decoded speech to determine a synthesis filter of a much higher order than conventionally used for decoding synthesis and then transmitting only the index for the vector which produces the lowest internal error signal. In other contexts, performing less critical portions of analysis on a previous frame of data has been termed "backward-adaptive". Two types of gain adaptation are used in respective implementations, both of which are backward-adaptive and one of which is dynamic to the extent that weighting coefficients are iteration-dependent. The spectral analysis is twentieth-order to fiftieth-order linear prediction analysis. Pseudo-Gray coding is used to reduce errors induced by channel noise; and the effects of the gain and specrtral analyses are taken into account in the training procedures which determine the sets of vectors which are the reference or candidate excitation vectors. |
