I was approached recently by IVO software, a leading TTS (Text To Speech) company, with their IVONA TTS service. Now on first sight you may think TTS is a rather old technology, which generally sounds like a disjoint Stephen Hawkings, struggles with acronyms and the contextual pronunciation of many words. But this is where IVO has achieved something quite special that caught my attention: an affordable, quality TTS service (runs in the cloud) using speech synthesis algorithms based on the elements of artificial intelligence and human speech as the interface in communication with a computer.
The other factor that stimulated my interest is market timing, given the growing challenge of accessing the web on smartphones. When I have a few spare minutes, while on the road, being able to review friends’ weblogs and articles I’ve bookmarked on my smartphone is very convenient. But to be frank it’s not that easy on a smartphone given the screen size (even on my iPhone 4G with its ‘retina’ display). This brings me back to IVONA, their impressive TTS enables me to listen to the web without the hassle of creating and managing podcasts. Its simplicity makes it just so compelling, and you can even download the mp3 from the widget.
The process of converting text to speech is not straightforward. Texts are full of heteronyms (words that are written identically but have different pronunciations), numbers, and abbreviations that all require expansion into a phonetic representation. There are many spellings in English which are pronounced differently based on context. For example, “My latest project is to learn how to better project my voice” contains two pronunciations of “project”.
IVO caught my eye, I’ve added their service to my weblog, see the Listen button (or play icon) at the top of the screen above the ‘Recent Comments’ section. I hope it makes this weblog more accessible given the reality of how we access the web these days. Try it out, and let me know what you think, thanks. UPDATE: In the emergency move to WordPress, done in April 2013, I’ve not yet re-enabled this function.