ModelTalker Interactive Demo
Below is an interactive text-to-speech form**
that demonstrates ModelTalker with different talkers (some
professional). Most demonstration voices are hybrid DNN (hDNN)
synthesis made with standard 1600 sentence inventories. However, check
out the several Tom voices that illustrate quality differences for
inventories of 400, 1600, and 3155 sentences using both unit selection
(uSEL) and hDNN technology.
Please type or paste some sample text below (one sentence at a time):