In the not-too-distant future, when a superstar artist like Drake releases a new song, we may be able to hear it in multiple languages, in the artist’s own voice, on release day.
On Monday (April 15th), we got a glimpse of how AI vocal technology can be used to do just that.
South Korea-based entertainment giant HYBE releases new single ” masquerade Written by the artist MIDNATT, the song is promoted by HYBE as “the first multilingual track produced in Korean, English, Japanese, Chinese, Spanish and Vietnamese”.
According to HYBE, MIDNATT sang the track in six languages and used AI to “apply native-speaker pronunciation data to the track to further refine the artist’s pronunciation and intonation.”
HYBE explains that the multilingual truck features technology from fake voice AI company Supertone, which it acquired last year in a deal worth around $32 million following an initial investment in the startup in February 2021. there is
According to HYBE, the project was launched “with the aim of providing music fans around the world with a music experience that transcends the physical limits of K-pop.”
In addition to the song being sung in six different languages, HYBE says they used voice design techniques to create the “artificial female voice” that can be heard on television. masquerade.
According to HYBE, “MIDNATT needed diverse vocals to intuitively express the self-contradictory emotions in the song, but BIGHIT MUSIC and HYBE IM used Supertone technology to create his own unique sound. I created a new voice while maintaining the vocal style.”
The company added: masquerade It mimics the effect of several singers singing on the track. ”
You can listen to a multilingual teaser for the track below, and listen to the full English and Spanish version of the track below.
MIDNATT is the alter ego of Lee Hyun, an artist signed to BIGHIT MUSIC.
MIDNATT (Lee Hyun) is the artist behind ‘Project L’, a collaboration project between HYBE’s ‘HYBE IM’ and BIGHIT MUSIC divisions, and HYBE’s chairman Bang Si-hyuk (aka ‘Hitman’ Bang). With the cover story mentioned in the April article signboard.
masquerade It is described as “a synthwave track with a modern take on a retro sound layered with rhythmic electric guitars”.
K-Pop’s “mega producer” and DJ Hitchhiker, who produced EXO, Red Velvet, f(x), etc., participated in the production of the track as an executive producer.
To celebrate the release of the single, MIDNATT held a media showcase with HYBE IM president Wooyoung Jung and BIGHIT MUSIC president Yeon Jaeshin.
“A lot of new technology is used in this project, but at the heart of the music is my story as an artist and as a human being.”
mid nut
MIDNATT said: “This project uses a lot of new technology, but at the core of the music is my story as an artist and as a human being.
“I received a lot of love in my previous job and it has been my motivation for a long time and I am very grateful.
“However, I had a yearning to try out a musical style that I had never tried before.
Founded in 2020, Supertone claims to be able to “create highly realistic and expressive voices.” [is not] Distinguishable from real humans. ”
The startup received global media attention in January 2021 for its so-called Synthetic Singing (SVS) technology.
MBW said in October that a major foray into artificial intelligence could be next on the agenda as HYBE ramps up its “diversification strategy to reduce its economic reliance on flagship BTS.” asked if
The use of AI vocals in the music business has been a hot topic ever since, and the new release of HYBE from MIDNATT shows what’s possible with Supertone’s technology.
Being able to release six versions of the same song in different languages on the same day to streaming services around the world, all in the same quality, with a real artist singing the song is a huge opportunity for rights holders.
In the wake of the Fake Drake story and other examples of AI software being used to imitate the voices of prominent artists without their consent, there has been a growing interest in generative AI and AI vocal mimicry technology among parts of the music industry. It’s not surprising that there is a lot of suspicion about its use.
But the latest efforts by Supertone and HYBE highlight that it’s not all threats. Today’s news follows suggestions from various high-profile music executives in recent weeks that AI could potentially have a positive effect on the music industry.
For example, Warner Music Group CEO Robert Kinkle told analysts at WMG’s earnings call last week, “When it comes to generative AI, we need to put it in the right context. It’s inaccurate to just view it as a threat. He added, “Our first priority is to enforce copyright and name, image, likeness and voice rights strongly to protect the originality of artists and songwriters.”
At Universal Music Group’s first-quarter earnings call, UMG chairman and CEO Sir Lucien Grainge said one day UMG will license some of its sound recordings to companies like OpenAI (creators of ChatGPT). You gave an interesting answer to the question whether you would consider it. Use a platform like this to legally build new music based on “derivative” copyrights. (Think about it: Millions of kids are making their own “Fake Drake” rap songs, but legally, and Drake gets paid for it.)
“We are open to any business solution when it comes to licensing,” Grainge said. “clearly [within that]must respect the integrity of the artists and their work.
“My philosophy for this company has always been: [that] We should and can be the “most hostesses”.
“Yes, we are open for business.” [AI companies] Which is legal and which is supportive, [with] That way, we can build partnerships for growth. ”global music business
