server.mrcp1.osrspeechrecog.startOfSpeechOnDTMF
                                                    Send a START-OF-SPEECH event on DTMF input.
| 
                                                                     Value  | 
                                                                
                                                                     Boolean DEFAULT: 1 (enabled)  | 
                                                            
| 
                                                                     How to set  | 
                                                                
                                                                     If using the Management Station, set on the Speech Server service. If not using Management Station, set in an NSS configuration file (user-NSSxx.txt).  | 
                                                            
| 
                                                                     Usage  | 
                                                                
                                                                     Occasionally changed at initialization. External integrators use this parameter. Frequency depends on customer: always or never. Applicable for Nuance Recognizer. Not applicable for Dragon Voice recognition.  | 
                                                            
This enables the voice platform to detect DTMF barge-in.
When a caller speaks, the server can send a START-OF-SPEECH event to the MRCP client. The client can use this information to detect that the caller has interrupted a prompt (known as barge-in). If this parameter is set to 1, Speech Server also sends a START-OF-SPEECH event when a touchtone occurs.
Although this parameter is specific to speech recognition, it relates to barge-in that occurs during a text-to-speech prompt: most MRCP clients send a BARGE-IN-OCCURED message to the server after receiving the START-OF-SPEECH event.
Supported for MRCPv1 browsers. (Ignored for MRCPv2 browsers.)