Hello,
Depending on the level of the sound signal that you will be receiving, doing an analog input and then in software determining if the level if your signal is high enough would be your best option if the sound signal you are receiving is not a TTL signal. But, as far as then generating a digital output that then triggers the analog input externally doesn't seem to be needed. If you are producing a software signal to trigger something anyway, you may as well just start the analog acquisition in software once your subvi has determined it is time instead of generating a digital output that then hardware triggers the acquisition.
Regards,
Steven B.
National Instruments