Digital Gating of Speech Signals

Abstract

Temporal gating of speech signals by means of a digital computer offers many advantages over analogue gating techniques currently in use. Prior to digital processing the analogue speech signal has to be converted into its digital representation. Requirements on sampling rate, number of quantization levels, and frequency response of the smoothing filter which are needed for distortion-free digitization and reconstruction of the analogue speech waveform are presented. Features of a gating program currently in use in this department are described.