Jump to content

AI or not AI?


Bapu

Recommended Posts

Macca says "The Beatles" will release a final song that was a demo from the Let It Be sesions where they got a "pristine Lennon vocal" and some George work done too.

I recall Peter Jackson saying that they used a tool they developed to reduce the guitars plucking during a convo between George and John and some background noise in the canteen with the lads having a deep conversation.

Your opinion? Is really AI or is it just more sophisticated filtering/EQ/Compression tools not yet available to us mere mortals? I'm reminded of tools like Moises (albeit maybe more primitive based on Jackson's description) where  you can extract vocal, drums, bas and "all other instruments" from a stereo recording.

Link to comment
Share on other sites

4 hours ago, Bapu said:

Your opinion?

Creative phase manipulation?

There is a plug in Ozone 9 Advanced that allows you to adjust the level of some instruments on a mixed down track. I used it quite a bit on my last project on old 4 track tape  tranfers. I rike it a rot.

Link to comment
Share on other sites

13 hours ago, Bapu said:

Macca says "The Beatles" will release a final song that was a demo from the Let It Be sesions where they got a "pristine Lennon vocal" and some George work done too.

I recall Peter Jackson saying that they used a tool they developed to reduce the guitars plucking during a convo between George and John and some background noise in the canteen with the lads having a deep conversation.

Your opinion? Is really AI or is it just more sophisticated filtering/EQ/Compression tools not yet available to us mere mortals? I'm reminded of tools like Moises (albeit maybe more primitive based on Jackson's description) where  you can extract vocal, drums, bas and "all other instruments" from a stereo recording.

https://www.theguardian.com/music/2023/jun/13/ai-used-to-create-new-and-final-beatles-song-says-paul-mccartney

The idea to use AI to reconstruct the demo came from Peter Jackson’s eight-hour epic, Get Back. For the documentary, dialogue editor Emile de la Rey used custom-made AI to recognise the Beatles’ voices and separate them from background noise.

A bit more here

https://arstechnica.com/information-technology/2023/06/ai-powered-final-beatles-song-due-this-year-says-paul-mccartney/

During the making of the 2021 Get Back documentary, composer Emile de la Rey and a team of engineers at Peter Jackson's Wingnut Films worked with machine learning researcher Paris Smaragdis of the University of Illinois Urbana-Champaign to develop a neural network called "MAL" (Machine Assisted Learning, also a nod to Beatles roadie Mal Evans) to isolate voices and musical instruments in Beatles recordings. Similar vocal isolation allowed McCartney to virtually duet with Lennon on a recent tour.

Traditionally, when multiple voices and instruments are recorded together on the same tape, it's difficult or impossible to isolate them without significantly losing sound quality or picking up frequencies from other instruments present in the recording. With the new technology, machine learning techniques allowed MAL to identify and pluck out individual audio elements, re-synthesizing them in a realistic way that matched trained samples of those instruments or voices in isolation. This allowed for boosting voices in noisy situations and also enabled surround sound remixes of the performances in the film.

But yes I get your point @Bapu .. we say every "new" thing like this to be AI when in fact it is not necessarily so

Edited by aidan o driscoll
Link to comment
Share on other sites

From what I can tell, they used Spleeter (which is AI based) to separate John's voice from the rest of the music, but beyond that there's no AI.

It reads like, "It's not John singing, it's AI" - but that's not the case.  It is John singing - they just used AI to get rid of the other instruments.

It's really not that much different from Queen's "Heaven for Everyone",  which used Freddie's vocal track from the same song on The Cross's "Shove It" album.  I've listened to both versions, and the vocal performance is identical.

 

  • Like 2
Link to comment
Share on other sites

27 minutes ago, msmcleod said:

From what I can tell, they used Spleeter (which is AI based) to separate John's voice from the rest of the music, but beyond that there's no AI.

It reads like, "It's not John singing, it's AI" - but that's not the case.  It is John singing - they just used AI to get rid of the other instruments.

It's really not that much different from Queen's "Heaven for Everyone",  which used Freddie's vocal track from the same song on The Cross's "Shove It" album.  I've listened to both versions, and the vocal performance is identical.

 

 

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...