I have long been using Windows Text to Speech (TTS) for several years now. I've favored the Win solution over Mac, given the poor implementation that had been available under Mac OS X (or 9 for that matter). The speaking voices were poor, the CPU usage was heavy and the time to rip TTS was too cumbersome. It's not easy for me to concede to the Win platform but I'm not that dogmatic about my choice of tools.

Today, I stumbled upon Cepstral quite by accident and found myself with a solution that now surpasses my Windows TTS solution (which was using TextAloud, a competent TTS application and NeoSpeech Voices) by way of rip speed. The Cepstral voices are on par with NeoSpeech but rip times are remarkably quicker. I don't know if this has anything to do with Apple's new Core services, but I like the result. I'm currently using Books2Burn for TTS to aiff but I expect I'll exploit OS X's more robust scripting interface to roll my own workflow.

Cepstral integrates seamlessly into Apple's TTS via the System Preferences panel. The demo has an appropriately annoying reminder to buy the full version but it is sufficient for your personal review before you select the best voice. Pricing is the same on all platforms (and on par with Window's commercial voice packages), which I find fair (especially since commercial developers typically penalize Mac ports with higher costs for the same ware.)