Author Topic: "How Audio Pros ‘Upmix’ Vintage Tracks and Give Them New Life" — Wired  (Read 548 times)

kimhill

  • Member
  • ****
  • Posts: 421
    • Instagram
"Using machine learning, engineers have made inroads into “demixing” the voices and instruments on recordings into completely separate component tracks, often known as stems. Isolating the components of songs is a surprisingly hard problem—more like unswirling paint than using a pair of scissors. But once engineers have stems, they can take the isolated tracks and “upmix” them into something new and perhaps improved. They might enhance a muffled drum track on an old recording, produce an a capella version of a song, or do the opposite and remove a song’s vocals so it can be used as background in a TV show or movie."

https://www.wired.com/story/upmixing-audio-recordings-artificial-intelligence/

Retronic

  • Administrator
  • *****
  • Posts: 2218
I find it appealing and imagine it a great way to pass a bit of free time.  I keep an eye out for a deal on a full Izotype RX to get my hands on the music rebalance program.  I'm more of a dabbler, mixing music for fun so it's a bit expensive for that but price will drop. 

stackjackson

  • Administrator
  • *****
  • Posts: 2198
    • GrooveRevolution
Fascinating. I had no idea this has been going on for the past 20 years. I'm so out of it.

The new library...

"high-quality stems ready for licensing in film, broadcasting, video games, and elsewhere"
« Last Edit: April 18, 2021, 03:28:49 PM by stackjackson »
| Stack |

Upgrade

  • Member
  • **
  • Posts: 61
I always found the effect interesting like those disembodied lips singing only the vocal tracks of famous songs in the Amazon Music commercial. Though some of the more modern songs in the other variants may still have separate tracks recorded.

youtube.com/watch?v=jbS6jzrD8G4

A lot of library music continued to be recorded in mono until the 1970s. It would be interesting if any labels decide if they want to spruce up their catalog or continue to promote them as vintage sounding recordings.

Porn Library

  • Member
  • ****
  • Posts: 374
"or do the opposite and remove a song’s vocals so it can be used as background in a TV show or movie."

https://www.wired.com/story/upmixing-audio-recordings-artificial-intelligence/

This has been going on since the 70's (at least in American library music). Many US library cuts from back when were either unreleased vocal tunes transformed into library background music, or private press released vocal tunes remixed into library background music. Most of Roland Hanneman's (R.S. Hanneman) Valentino stuff was this, as well as some of the more obscure hard to find US library like Dick Hieronymus and Don Great material.

potzorbie

  • Member
  • **
  • Posts: 86
I always found the effect interesting like those disembodied lips singing only the vocal tracks of famous songs in the Amazon Music commercial. Though some of the more modern songs in the other variants may still have separate tracks recorded.

youtube.com/watch?v=jbS6jzrD8G4

A lot of library music continued to be recorded in mono until the 1970s. It would be interesting if any labels decide if they want to spruce up their catalog or continue to promote them as vintage sounding recordings.
That Whitney Houston vocal used in that commercial is from a pre-existing acapella on the 12" single released back when the song came out.

kimhill

  • Member
  • ****
  • Posts: 421
    • Instagram
One thing I'd like - if it were easier to fix early stereo mixes. Back when stereo was new, a lot of engineers had instruments panned hard left or hard right, which isn't so great with headphones.