Abstract
An important information that is continuously being transmitted from the analyser to the synthesiser of a channel Vocoder, concerns the nature of speech sound and pitch. Speech sounds are complex in nature and consist of voiced as well as unvoiced sounds. In the analyser a decision is first made whether the sound is voiced or unvoiced. In case it is voiced then a signal is obtained proportional in magnitude to the pitch. The quality of the synthesised speech depends to a considerable extent on this information. This paper discusses the acoustical and circuit consideration on which the design of the voicing unvoicing module for the 16-Channel CEERI Vocoder is based.