Sunday, January 09, 2005
1. By the Speech SDK (also called SAPI) we can convert text to a WAV file. But there is no such Text-To-Speech (TTS) engine or SDK for .NET Compact Framework. In another word, we can not let smart device read the text for us.
So the only way under this architecture is to set up a web service to convert text to WAV files; the smart device clients send a request to the web service and play the returned WAV files. I already finished a program to play WAV on my HP ipaq. It works but I am not sure if the performance is good enough to transfer WAV files.
Usually WAV files are bigger than MP3. So another approach is to convert WAV files to MP3 format before transferring them. But I am not sure if I can play MP3 files on the pocket pc by some API.
2. The other choice is Speech Application SDK (also called SASDK), which is tied with ASP.NET.
Good news is there is a Pocket Internet Explorer add-in for smart devices. And this add-in provides all the functionalities we need such as TTS even voice recognition. http://msdn.microsoft.com/library/default.asp?url=/library/en-us/sasdk_usermanual/html/UM_bsa_Client_IE.asp
Bad news is this approach needs an extra speech server http://www.microsoft.com/speech/evaluation/tools/ . And at the first glance, this server does the same thing we try to implement! But in a more graceful way: wrapped in XML.
Based on the above info I can get so far, I think it is better to focus on how to retrieve/organize info from the web pages according to some user-specified profile.