1.2.5 Recording or Streaming to the Host

You can choose to record (stream) data at all times or only after the VAD has detected speech. Voice-activated recording refers to the process of recording or streaming data to the host application only after the VAD has detected speech. Data in the pre-speech buffer is also sent to the host application as part of this process.

Note:

In the industry, the terms recording and streaming are often used interchangeably. In this document, the terms have slight differences in meaning. Streaming refers to the transfer of data from the firmware to the application on the host. Recording refers to the transfer of data from the firmware to a file or memory buffer on the host.

The ECCH_VADINITIATED parameter in ec_setparm( ) controls whether speech data is transmitted constantly or only after the VAD detects an utterance. Data is transmitted using the ec_reciottdata( ) or ec_stream( ) functions. For more information, see Section5.1.4, Sample VAD Scenarios and the function descriptions in the Continuous Speech Processing API Library Reference.

Click here to contact Dialogic Customer Engineering