Perceptual Domain Based Speech and Audio Coder

This paper applies a new auditory filterbank to wide band speech and audio coding. The coding algorithm is capable of producing high quality coded speech and audio, which account for temporal as well as spectral details. The analysis and synthesis are performed using a critical-bandrate auditory filterbank with superior auditory masking properties. The outputs of the analysis filters are processed to obtain a series of pulse trains that represent neural firing. Post and pre temporal masking models are applied to reduce the number of pulses in order to produce a compact timefrequency parameterization. The pulse amplitudes and positions are then coded using run-length coding algorithm.