Re:Audio's Circle of Confusion (2009)
by actionfromafar on Thursday April 13, 2023 @07:26PM
(#35563558)
Something like this could definitely be a thing, with AI or some kind of physical simulation or both.
Imagine some Jazz recordings by some old band. There happen to be a few in high quality LP? and some in low quality (shellac?). One could imagine some kind of programming magic which transplanted only the "feel" and "timbre" of the hi quality recording onto the low quality recordings.
Further:
The simulator could build up full instruments and musicians in a studio and mike them up and modify them until the virtual simulated chain's end result is the same mix as the LP. That's a baseline. Now, you can isolate all the tracks and remix the album as if every musician was miked up separately from the beginning, even though the originally played live with a mono microphone.
It wouldn't be real, but it would be a perfectly plausable version and very close to real, too. Of course, the more you have to extrapolate and transplant, the larger the risk of ending up like that famous "enhanced" photo of Obama, but in the audio domain.
Imagine some Jazz recordings by some old band. There happen to be a few in high quality LP? and some in low quality (shellac?). One could imagine some kind of programming magic which transplanted only the "feel" and "timbre" of the hi quality recording onto the low quality recordings.
Further:
The simulator could build up full instruments and musicians in a studio and mike them up and modify them until the virtual simulated chain's end result is the same mix as the LP. That's a baseline. Now, you can isolate all the tracks and remix the album as if every musician was miked up separately from the beginning, even though the originally played live with a mono microphone.
It wouldn't be real, but it would be a perfectly plausable version and very close to real, too. Of course, the more you have to extrapolate and transplant, the larger the risk of ending up like that famous "enhanced" photo of Obama, but in the audio domain.
The same could be done for video, too.