This definitely seems like synthetic speech. I would bet that it is AT&T Naturally Speaking or something like it which uses a similar technology to perform text-to-speech based on sampling from a human subject. I recently heard voices with similar quality from the samples provided on the Freedom Scientific web site for their Sarat reading system. Quite amazing - Too bad this can't be done in a small unit like the Bookport with minimal processing power. This type of speech reproduction probably requires a fair amount of horsepower. Some day, we'll be there though! -- Pete