Parasite
Newbie
- Joined
- Jul 3, 2003
- Messages
- 377
- Reaction score
- 0
Well if phenome recognition was built into the compression codec or whatever it could work, even if crudly. It wouldnt need to recognize every phenome, so long as it could differenciate hard sounds from soft sounds and maybe a few hard vowel sounds, and of course syllables and silence (ie no words)...For example t, sh, ch, th, k, j, could all work with the same lip formation, just as p and m could, or o, q, and , as well as some other combinations that would not get confused, then of course a few unique lip movements like f, "hard a", "hard e" and "hard o"...All in all I think you could get away with less than 10 lip formations and a simple form of syllable/phenome recognition.
But hell I would be fine with open mouth/closed mouth if it were accurate to the syllable.
But hell I would be fine with open mouth/closed mouth if it were accurate to the syllable.