3 ms·

> Unfortunately, most of this material was only released in unofficial forms, often on low-quality MP3s I know nothing about audio - is it possible to upscale

> Unfortunately, most of this material was only released in unofficial forms, often on low-quality MP3s I know nothing about audio - is it possible to upscale the quality somehow? Improve the quality via some process?

jl6 3y ago
Technically no, pragmatically maybe… If the information has been lost through using a lossy codec like MP3, there’s no getting it back. But there are “AI upscaling” techniques that use models to fill in the missing pieces. It isn’t really restoring the original, more like painting over a crack in a style that is typically used by similar artists in similar contexts. Such models could produce a result conceptually similar to what a cover band might produce.
giantrobot 3y ago
Audio compression literally throws out data. It's audio the human ear isn't likely/able to hear but it's still gone. Different codecs and even different implementations of codecs throw out different audio data. Any sort of upscaling is going to involve a lot of guesses about what was thrown out. For every codec and implementation there will be a lot of possibilities of what the original data might have looked like. At the end of the process you likely aren't going to get a result that necessarily sounds better than the low quality encoding. There will likely be a lot of artifacts introduced by the upscaling. It's not impossible but something you'd need a specially trained AI to do I would think.
- lostlogin 3y ago
  I work with a system which extrapolates data in the frequency domain to reduce noise in medical images and then doubles the images resolution in the x and y direction (so 4x the pixels). It’s absolutely amazing. It’ll be interesting to see if audio ever gets a similar treatment and in what areas it can be applied.
  - PaulDavisThe1st 3y ago
    Noise levels in contemporary audio is vanishingly small, and has not really been the target of any audio technology for quite a long time. Of course, restoration of older recordings is a separate issue, and that technology is 20-30 years old, at least (and very, very good).

jl6 3y ago

Technically no, pragmatically maybe… If the information has been lost through using a lossy codec like MP3, there’s no getting it back. But there are “AI upscaling” techniques that use models to fill in the missing pieces. It isn’t really restoring the original, more like painting over a crack in a style that is typically used by similar artists in similar contexts. Such models could produce a result conceptually similar to what a cover band might produce.

giantrobot 3y ago

Audio compression literally throws out data. It's audio the human ear isn't likely/able to hear but it's still gone. Different codecs and even different implementations of codecs throw out different audio data. Any sort of upscaling is going to involve a lot of guesses about what was thrown out. For every codec and implementation there will be a lot of possibilities of what the original data might have looked like. At the end of the process you likely aren't going to get a result that necessarily sounds better than the low quality encoding. There will likely be a lot of artifacts introduced by the upscaling. It's not impossible but something you'd need a specially trained AI to do I would think.

lostlogin 3y ago
I work with a system which extrapolates data in the frequency domain to reduce noise in medical images and then doubles the images resolution in the x and y direction (so 4x the pixels). It’s absolutely amazing. It’ll be interesting to see if audio ever gets a similar treatment and in what areas it can be applied.
- PaulDavisThe1st 3y ago
  Noise levels in contemporary audio is vanishingly small, and has not really been the target of any audio technology for quite a long time. Of course, restoration of older recordings is a separate issue, and that technology is 20-30 years old, at least (and very, very good).