There was a bit of FUD surrounding the decision to use a form of AI to resurrect John Lennon’s voice on what Paul McCartney called “the last Beatles record.” What they’ve done is a far cry from the sketchy AI imitations of artists we see crashing Soundcloud today, and has a lot more in common with a more prosaic application of machine learning: noise reduction.
To hear people talk about it, you’d think it was an outrageous cash grab using the latest voice-synthesis technology to ape one of the most famous songwriters of all time. But the real story is simpler and more painful than that, and the technology is far less fantastic.
As the band members recall in a sweet short film about the making of the song, “Now And Then” was originally a piano demo Lennon made shortly before he was killed in 1980. His widow, Yoko Ono, gave the tape in which It was recorded in the band afterwards, but the quality of the recording was not great — bad, even.
“When we were listening to ‘Now And Then,’ it was very difficult because John was kind of hidden,” Ringo Starr says in the short.
“Every time I wanted a little more of John’s voice,” McCartney recalls. “That piano came in and blurred the picture. And in those days, of course, we didn’t have the technology to do the separation.”
In 1995 they “ran out” when they tried to save the song, but in 2022 they were working with Peter Jackson on the documentary “Get Back”. The director and his team applied modern sound processing technology to archival footage of the band to isolate individual instruments and vocals.
“We paid a lot of attention to the technical restoration. This ultimately led us to develop a technology that allows us to take any soundtrack and split all the different elements into separate pieces based on machine learning,” Jackson said in the brief.
MAL, as they called it, is a version of sound isolation technologies that have come a long way in recent years. Machine learning models can be trained on, say, multiple guitar tracks and learn what the waveform or spectral signature of a guitar is, and they can with varying success extract it directly from a mixed track.
It’s also commonly used in video calls now, using models trained on human voices. By suppressing anything that isn’t the speaker’s voice, background noise like dogs barking or a strong coffee can be silenced in real time. Harder versions of this were sometimes used to make karaoke versions of songs by identifying and removing the vocal parts.
In the case of Lennon’s demo, it worked like a charm, as you can hear on this timestamp in the construction of the short.
“There it was—John’s voice, crystal clear,” Paul said. “Now we could mix it up and make it a proper record.”
Some may question the ethics of making this record, but everyone involved seems to believe that John would be all for it, as he loved tinkering with technology and, of course, had written and performed the song originally with the intention of record it.
But more importantly, it seems to have served as a little closure for the team. The vicissitudes of stardom and creativity they endured are more than well-documented, but to lose a friend and creative partner of decades in this way, and to have that last, lingering loose end dangling just a little too far, must have been excruciating.
As anyone who has lost someone can attest, every remnant becomes precious. “To hear John’s voice … that’s something to love,” George Harrison said in 1995.
And now, with a quarter of a century of technological improvements, they could do just that.
“He was the closest we’ll get to having him back in the room,” Ringo said.
You can listen to “Now And After” on the right here.
