It should be pretty easy to piggyback asynchronous push-to-speak voice chat over IRC or another existing chatroom system. One speech compresses a recording and transmits ASCII encoded. Tones corresponding to a partial hash of the username can identify the speaker. Those without a microphone can participate via speech synthesizer either at the source or destination... an interesting exercise might be to synthesize directly to a speech compression format. The speakers should be turned off when recording. One can click to skip an utterance. There is a text side channel for URLs, spellings.
No comments :
Post a Comment