Paul McCartney will use artificial intelligence to rescue John Lennon’s voice on the new Beatles album

One of the last members of the Beatles, Paul McCartney recently announced that what he calls the British group’s latest album is in production, for which he is using artificial intelligence to extract the voice of John Lennon from an old home recording.

In an interview with BBC Radio 4McCartney indicated that this method was used as the original song would have been recorded in the year 1978 by lennon in his apartment in New York (so it did not have studio quality) and it was not finished, but with the technology current could be launched during the year 2023.

The song, which the British outlet indicates could be called “Now and Then” would be part of a series of recordings made in cassettes intended for McCartney and that would have been compiled shortly before the murder of the musician in 1980. His partner, yoko onowas in charge of sending the package to the former Beatles.

According to BBC This would not be the first time that an attempt would be made to work on “Now and Then”, as the producer Jeff Lynne, who was also in charge of other songs by the group released after Lennon’s death, indicated that between 1995 and 1996 the band members tried to record the song, but “the song had a chorus, but it had almost no verse. We did a fast track that we never really finished.”

McCartney stated that George Harrison, who died in 2001, refused to work on that song further because he considered the quality of Lennon’s recording to be lousy. The problem would be that in the Audio from the singer a constant hum was heard as a result of the installation of electrical circuits in his apartment.

The idea that led to the launch of “Now and Then” arose during the production of the documentary Get Back, available at Disney+ and in which the story of the album is told let it be. In this process, computers trained to differentiate the voices of the band members were used to then digitally separate them into clean audio tracks without instruments.

Using this process, the director of the documentary, Peter Jackson, managed to extract Lennon’s voice. “They told the machine: “This is the voice. This is the guitar. Eliminate the guitar. That way it was only necessary to work on the mix of the song as you normally would. “That gives you some freedom,” McCartney told BBC.

In the same way, this technology allowed the musician to perform a duet with lennon during one of his most recent tours in United Kingdomin addition to facilitating the reissue of the album Stir with surround sound that was released in November 2022.

Unlike cases in which fans of some artists take recordings of voices and generate new verses, as well as musical tracks using artificial intelligence, the process through which the recording of lennon it is only to differentiate the musical background and the singer’s voice in two tracks, so it can continue to be considered that it is he who sings and not a machine trained to imitate him.

for his part McCartney He indicated that even though he is scared by the possibility that an artificial intelligence could imitate a voice and sing, “it is exciting because it is the future. We’ll have to see where it takes us.”