|
|
|
|
|
by parkan
5178 days ago
|
|
Rebuilding speech/song based on existing phoneme library (i.e. the discography) is already practical and, like everything else in this domain, has been done long ago in japan: http://www.wired.com/underwire/2011/12/ueki-loid-speech-synt... See also Hatsune Miku/Vocaloid, which is a fully artificial pop star that's been selling out shows (as a hologram since 2009) |
|
It would be in such bad taste that even the American recording industry wouldn't go near it. People already get upset that they use old entire recordings to produce new songs, and those are actually his lyrics and his voice.