| Platform SDK: Agent |
The recommended system configuration for using the sound editor is a PC with a Pentium 166, at least 48 MB RAM, and a Windows-compatible sound card. If you want to record spoken input with the tool, you will also need a compatible microphone.
To install the Microsoft Linguistic Sound Editing Tool, open its self-extracting installation file. This will automatically install the appropriate files on your system. If you download the sound editor from the Microsoft Agent Web site, you can choose to install the editor after downloading or save it to your disk to be subsequently opened and installed. The installation tool will propose to install itself in the Tools subdirectory of Microsoft Agent. We recommend that you use this location.
The Microsoft Command and Control speech recognition engine (version. 4.0) must also be installed before you can use the sound editor. This normally gets installed with the sound editor, but if it was subsequently uninstalled, you can reinstall it from the Microsoft Platform SDK or the Microsoft Agent Web site at http://www.microsoft.com/msagent/downloads.htm#sr. The sound editor can only generate linguistic information based on the language supported by the speech engine. To generate information for other languages, a compatible speech recognition engine for that language must be installed. Contact your speech engine vendor to determine whether they support the Microsoft Linguistic Sound Editing Tool.