Configuring Speech Engine Services

  • Article

Configuring Speech Engine Services

Use the Speech Services Settings tab of the Microsoft Management Console (MMC) snap-in for Microsoft Speech Server (MSS) to configure Speech Engine Services (SES) settings.

For instructions on how to open the MMC snap-in for MSS, see Administering Speech Server Using MMC.

To configure SES settings

  1. Open the MMC snap-in for MSS, and click the Speech Services Settings tab.

  2. Under Cache settings, in Preloaded resource manifest URL, type the path to the application manifest file that SES retrieves at startup. This setting corresponds to the PreloadedResourceManifest Windows Management Instrumentation (WMI) property.

    The application manifest file lists grammar files and prompt databases that MSS should preload on SES startup for use by an application. See About Application Manifest Files.

    In a multiple application environment, the application manifest file should always be SESManifest.aspx, which is located on the Web server. SESManifest.aspx is dynamically generated and contains the preload list for all applications registered with the Speech Application Deployment Service (SADS). For more information, see Configuring SES and TAS for Telephony Applications.

  3. In Cache maximum size, type the maximum size of the disk cache that is used for storing prompt databases. This setting corresponds to the MaxDiskCacheSize WMI property.

  4. In Cache location, type the path to the folder where prompt databases are cached. This setting corresponds to the DiskCacheLocation WMI property. If the path is not specified, %TEMP%\CachedPromptDatabases is used.

  5. Under Audio settings, in Audio connection port, type the port number that SES monitors for audio connection requests from a client. This setting corresponds to the AudioConnectionPort WMI property.

    Note The active and idle worker processes must listen on different ports. Therefore, when the failover worker process is enabled (default), SES also uses the port number that follows the value you enter here. If you are running a firewall, you must open both ports for SES to function reliably.

  6. In Audio connection IP, type the IP address that SES monitors for audio connection requests from a client. This setting cCorresponds to the AudioConnectionIP WMI property.

    Select Use default to have SES use the first IP address on the first adapter on your server. This is the simplest choice if you have only one network card on your server.

    Select Use specific, and type the IP address you want SES to use, if your server has multiple network adapters, or if you are using Network Load Balancing (NLB). (With NLB, use the host IP rather than the cluster IP.)

  7. Select or clear the Enable audio compression check box as needed. This setting corresponds to the EnableAudioCompression WMI property.

  8. In Shutdown timeout (seconds), set the number of seconds allowed to finish all ongoing sessions. This setting corresponds to the ShutdownTimeout WMI property. Any sessions open at the end of this period are forcibly closed.

Remarks

The application reduces user wait times by preloading application resources. If resources are particularly large, memory consumption can become a problem. See Optimizing SES Memory Usage.

See Also

Configuring Speech Engines | Administering Speech Engine Services | About Application Manifest Files