what is a speech synthesis index?
first, why we don't use
signal processing
making the
index
into a database for resequencing
training the
weights
for selection