You could try the code in the following anvil thread:
I would suggest saving the audio files to Dropbox then reading them from there.