Analysis of speech waveform quantization methods

Abstract

Digitalization, consisting of sampling and quantization, is the first step in any digital signal processing algorithm. In most cases, the quantization is uniform. However, having knowledge of certain stochastic attributes of the signal (namely, the probability density function, or pdf), quantization can be made more efficient, in the sense of achieving a greater signal to quantization noise ratio. This means that narrower channel bandwidths are required for transmitting a signal of the same quality. Alternatively, if signal storage is of interest, rather than transmission, considerable savings in memory space can be made. This paper presents several available methods for speech signal pdf estimation, and quantizer optimization in the sense of minimizing the quantization error power.