Zenph Sound Innovation currently uses music analysis technology to create new recordings from deceased musicians. As reported this week in Wired, the process analyzes aged, damaged recordings for a musician's style, and creates a
virtual personality that can replay the piece just as the original person would have.
Zenph has already succeeded in creating new versions of previous pieces that are too distorted to appreciate, as if these musicians had access to contemporary recording technology. These new, clear recordings can be licensed to films, and software could let musicians play virtually with famous artists on a level way past Guitar Hero. In Buskirk's example, a guitarist could one day run his solo through an Eric Clapton program that would shape it to sound as if it were played by the latter.
Once the tech has evolved enough, it should be able to take an analyzed style, and feed a new composition through it, one that the original musician never played. In effect, a virtual artist could play any available song, whether he was alive or not when it was composed. The artist, or the estate of an artist, would give permission for the style, not the work, of a song to be licensed.
Additional concerns manifest when industry professionals try to define the new licensing terms. While musicians have been influenced by others for as long as the art existed, the imitation was done with a human brain. Now, style is being analyzed by a computer, and the machine must license that particular person's personality in order to do so.
Eric Singer, creator of the League of Electronic Musical Urban Robots (LEMUR) is not in favor of extending copyright, and even less in favor of extending copyright to style. “It basically means that the entire history of music, where people have listened to other musicians and been influenced by their style is basically up for grabs. Whether a brain is doing it or a computer is doing it, how are they going to make that distinction?”