Freedom Fone’s ability to fulfill it’s promise as a must have tool for bridging the digital divide has yet to be determined. Millions of poor people have access to mobile phones, but with tariffs as high as they are in countries like Zimbabwe, experimentation in this field is still costly. And of course, for our project these are early days. The development team is still in the process of creating the variety of features that will distinguish Freedom Fone from the technically intimidating (to ordinary folk) IVR products like FreePBX, Trixbox and PBX in a Flash.

One of the recalibrations for me has been a growing appreciation of the relevance of text-to-speech synthetic voices for our platform. This isn’t news to our Project Architect, Alberto Escudero Pascual. He’s been convinced of its relevance from the start. In fact, in order to build an interactive online demo for Freedom Fone he integrated a commercial synthetic voice from Cepstral called Allison as a quick option for building and testing a voice menu.

As you can imagine, English speaking Allison, as good as she sounds given she’s synthetic, is not an ideal voice for enunciating other languages.

As a project located in Africa we are keen to develop/acquire free synthetic voices for some of the continent’s many languages and include them with the Freedom Fone software. As an open source project I hope that we can attract the contribution of free synthetic voices for many of the world’s languages over time.

Yesterday I had the pleasure of speaking with Etienne Barnard at Meraka Institute in Pretoria, South Africa. To my delight he indicated that work already done in Kenya on text-to-speech for Kiswahili by a team led by Dr Mucemi Gakuru at the University of Nairobi some years ago, might be updated and made available in time for our July release of Freedom Fone version 2.

In recognition of the competitive mobile phone tariffs prevailing in east Africa and the willingness of organisations there to experiment with information on demand voice services, we will create our first localisation of the Freedom Fone GUI for Kiswahili in February 2010. The possibility of including a free synthetic voice for this audience is exciting.

So why this interest in synthetic voice? Doesn’t this just mean a horrible robotic sounding Kiswahili voice? Obviously original audio files with perfect inflection are the first choice, but not all information requires the effort associated with recording audio files. Freedom Fone helps with the automatic conversion of audio files for voice menus, and it will be improved over time to make it as easy as possible to create audio files using a basic microphone attached to a computer. Still, it would be a lot quicker to automatically convey information received/produced in text format, like product prices, weather reports, breaking news using text-to-speech.

And … not all synthetic voice sounds dreadful. Build and test your own voice menu in English using Allison and our online demo. Make it the default audio menu and call in to listen for free using Skype. To do this you will need to add Skypiax4 as a Skype contact. Let us know what you think of the experience!

This entry was posted on January 15th, 2010 at 12:22 pm by Brenda Burrell and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.