Low bit rate coding for speech and audio using mel linear predictive coding (MLPC) analysis

This paper proposes a low bit rate coding method for speech and audio using a new analysis method named MLPC (Mel-LPC analysis). In the MLPC analysis method a spectrum envelope is estimated on a melor bark-frequency scale, so as to improve the spectral resolution in the low frequency band. This analysis is accomplished with about two-fold increase in computation over the standard LPC analysis. Our coding algorithm using the MLPC analysis consists of five key parts: time frequency transformation, inverse filtering by the MLPC spectrum envelope, power normalization, perceptual weighting estimation, and the multi-stage vector quantization. In subjective experiments, we have investigated the performance of MLPC analysis method, through the evaluation of paired comparison tests between the MLPC analysis and the standard LPC one in inverse filtering. In all bit rates, almost all the listeners feel decoding signals by the MLPC analysis method is superior to the LPC one. Especially in low bit rate, there is a great difference between them.