Yeyep is Your Go-to Source for All the Latest Entertainment News, from Movies and TV Shows to Music and Fashion.
⎯ 《 Yeyep • Com 》

This K-pop artist uses AI to sing in 6 languages

2023-05-16 00:29
A new project from the company behind K-pop superstars BTS and New Jeans is using
This K-pop artist uses AI to sing in 6 languages

A new project from the company behind K-pop superstars BTS and New Jeans is using AI technology to sing in six languages. MIDNATT, a.k.a. the alter ego of 39-year-old singer Lee Hyun, debuted today with the single "Masquerade," which was originally recorded in Korean and engineered into English, Japanese, Chinese, Spanish, and Vietnamese with the help of AI audio tech company Supertone.

"MIDNATT sang the track in the six languages," says a press release, "and the pronunciation data of native speakers was applied to the track to further refine the artist’s pronunciation and intonation." Supertone was also used to create multiple artificial voices in the song, including that of a woman, using Lee's voice. "Masquerade" was produced by DJ Hitchhiker, who has made hits for artists under HYBE competitor SM Entertainment, including EXO and Red Velvet.

SEE ALSO: Is AI-generated music the future of K-pop? HYBE chairman Bang Si-Hyuk thinks so

The experiment has been called "Project L" internally and was mentioned by HYBE Chairman Bang Si-Hyuk in an April interview with Billboard (HYBE acquired Supertone earlier this year). The press release notes that the MIDNATT project was "launched with the goal of offering global music fans a music experience that goes beyond the physical limitations of K-pop."

The music video for "Masquerade" can also be "viewed" in six languages, but the video is exactly the same in each one. Only its audio changes. For that reason, you won't see MIDNATT mouthing lyrics in the video. We've embedded the video below so you can experience it for yourself. To change the language, click the settings button (white cog) in the lower right hand corner of the video player and select "audio track" from the menu to choose a language.