Current logic submit the data to tts streaming module to freeswitch once boundary is detected but it does not send flush command to start generating the audio which takes 2 sec additional to send flush based on timer value. it would good idea to send flush after detecting boundary there could be environment variable to control this feature
Current logic submit the data to tts streaming module to freeswitch once boundary is detected but it does not send flush command to start generating the audio which takes 2 sec additional to send flush based on timer value. it would good idea to send flush after detecting boundary there could be environment variable to control this feature