The instructions in the ssml file are read by TTS_callback. The markers in the ssml file are read by the supplementary code i have added to tts_callback and a cue.txt file is output to the cues folder. The cue.txt file contains a time in msecs, a file name (+trigger ‘extension number’ ‘io number’). The io numbers are 1 = receiver, 2 = kiosk, 3 = external, 4 = record from kiosk mic.

There will need to a limit on the number of trigger receivers available in PD and thus the number of trigger in any extension sequence.

<!-- the mark name format is 'filename channel' io channels are designated by numbers 1 = receiver, 2 = kiosk, 3 = external, 4 = kiosk mic
--> <?xml version="1.0" encoding="UTF-8"?>

This is an fx file embedded into an synthesis stream <audio src="/home/chris/Dropbox/K6/fx/100_1/100_1.wav"/>
This is audio file play routed to receiver.<mark name="triggerA 1;"/>
This is audio file play routed to kiosk. <mark name="triggerB 2;"/>
This is audio file play routed to external. <mark name="triggerC 3;"/>
This is audio file record from kiosk mic.<mark name="triggerD 4;"/>
This is record file play routed to receiver
This is record file play routed to kiosk.
This is record file play routed to external.
Now dial another number.
</xml>