I’d like to thank Liz Knights for her tremendous support, encouragement and enthusiasm; Virginia Bovell for her tolerance and understanding; Nick Coleman, Ian Craig, Ian Preece, Caroline Dawnay and Viv Redman
An improved cepstrum-based voicing detection and pitch determination algorithm is presented. Voicing decisions are made using a multifeature voiced/unvoiced classification algorithm based on statistical analysis of cepstral peak, zero-crossing rate,
The pitch calculation is based on the AMDF-fractional pitch period determination. The function is designed to operate optimally at the sampling rate of 8000Hz; nevertheless, you may adjust the window size (wlen) to suit your pitch watching needs.