An improved residual encoder for speech compression

An improved Residual Encoder is developed for digitization of speech for different languages such as: English, Chinese, Arabic, French, German and Russian, etc. Experiments show that the improved system achieves a 2-7dB increase in signal-to-noise ratio over the previous Residual Encoder (1). Effect of transmission bit error on the speech quality is also investigated, it is found that the variable input code of (1) suffers more degradation in speech quality than the fixed code used in this study with transmission errors. The implementation complexity is on the same order as other Residual Encoder Systems.