The sound files can be added to a certain sentence e.g.:
Code:
talk.s = "hello";
talk.file = "hello"; // file in '/ai/speech/character_name/hello.wav'
You could make this dependent on other states like this:
Code:
talk.s = "hello";
[state.dyn.me.coll.vag_deep.obj_par >= 1]
talk.file = "hello_highpitch";
else
talk.file = "hello";