This paper discusses the significance of segmental and prosodic knowle
dge sources for developing a text-to-speech system for Indian language
s. Acoustic parameters such as linear prediction coefficients, formant
s, pitch and gain are prestored for the basic speech sound units corre
sponding to the orthographic characters of Hindi. The parameters are c
oncatenated based on the input text. These parameters are modified by
stored knowledge sources corresponding to coarticulation, duration and
intonation. The coarticulation rules specify the pattern of joining t
he basic units. The duration rules modify the inherent duration of the
basic units based on the linguistic context in which the units occur.
The intonation rules specify the overall pitch contour for the uttera
nce (declination or rising contour), fall-rise patterns, resetting phe
nomena and inherent fundamental frequency of vowels. Appropriate pause
s between syntactic units are specified to enhance intelligibility and
naturalness.