Abstract
The present study investigated the effect of pitch and rate variations on the intelligibility of Echo-II produced speech. The results show that slowing down the rate of text-to-speech synthesized speech from the default value of about 201 syllables per minute to 139 syllables per minute enhances word intelligibility by more than 10 percent and message intelligibility by about 14 percent. High (194 Hz), medium (111 Hz), and low (82 Hz) pitch levels produced roughly equivalent intelligibility scores. Augmentative and alternative communication and educational applications of text-tospeech synthesis need to permit users the flexibility of varying rate of speech to enhance intelligibility. Clinicians and teachers also need to note that message intelligibility is still only 58 percent with a slow rate of speech.

This publication has 5 references indexed in Scilit: