First, I would like to point out that the Microsoft Speech Pack (Speech SDK 5.1) is very powerful, but I have only used a small portion of its features. Recently, I found out that there is already version 5.3, and last night, I accidentally discovered that dictionary tools like Lingo and Kingsoft have utilized its TTS (Text-to-Speech) component.
First of all, you must install the SPEECH SDK 5.1 along with its additional voice packs (Japanese and Chinese, while English is included in the SDK package). After installation, it's best to initialize the system and configure the settings. This involves simply recording your voice.