| 摘要 |
For voice and musical signal, a signal coding system which can obtain good sound quality even at a low bit rate. The signal coding system predicts an input signal in a predicting circuit and performs orthogonal transformation in a first orthogonal transformation circuit of a predicted residual error signal. In a coefficient calculating circuit, a coefficient of smaller degree is calculated for expressing an envelop of the orthogonal transformation coefficient in the orthogonal transformation circuit. In a quantization circuit, quantization is performed by expressing the orthogonal transformation coefficient with a plurality of pulse trains with determining positions for generating a pulse using the output of the coefficient calculating circuit. The envelop of the orthogonal transformation coefficient is expressed by the coefficient with smaller degree. On the basis of the coefficient, the orthogonal transformation coefficient is expressed by a plurality of pulse trains to perform more efficient coding than that of the prior art. <IMAGE> |
