Multimode Speech Coding - Digital Speech: Coding for Low Bit Rate Communication Systems - page 284

t(n)

e(n)

r(n)

0 100 200 300 400

samples

(a) Male offset

t(n)

e(n)

r(n)

0 50 100 150 200

samples

(b) Female onset

Figure 9.5 Irregular pulses at the onsets and offsets

further refinements are required. Moreover, the regions where e(n)>t(n) ,
gives only a crude estimation of the pitch pulse location. The algorithm relies
on the accuracy of the estimated pitch used for the computation of t(n) and
in the refinement process described below. However SWPM needs only the
pitch pulses in the stationary voiced segments, for which the pitch estimate
is reliable.