23 Apr
2012
23 Apr
'12
1:20 p.m.
Hi, I have a asr dialogue with two consecutive play_and_detect_speech calls. The say: part is empty and the detect: part uses nuance grxml grammar. In the second play_and_detect invocation freeswitch looses some of the userplane data, thus causing the speech detection to fail.
From the userplane data forwarded by freeswitch to the Nuance NSS server, about one second of speech is missing in the beginning of the phrase
If I add say: with some content, then all works fine. What could cause this loss of userplane data ? Cheers /Mike