Big Ocean, claimed to be the “world’s first deaf K-pop group”, released their new single using voice cloning technology from Supertone, HYBE’s AI audio subsidiary.
The track, called Bucket Hat, premiered at the UN’s AI for Good Global Summit in Geneva last week (July 9) before releasing on streaming platforms, The Korea Herald reported.
Supertone’s Controllable Voice Conversion system was used to process vocals from Big Ocean members PJ and Jiseok, extracting their vocal timbre and blending it with fluent English pronunciation from a guide vocal, according to the report. The technology allows the artists to preserve their distinctive sound while improving clarity for international audiences, the report said.
In a LinkedIn post on Monday (July 14), Supertone said: “We worked on enhancing their English pronunciation, preserving Big Ocean’s unique vocal tone and style as they share their music with a global audience.”
“We worked on enhancing their English pronunciation, preserving Big Ocean’s unique vocal tone and style as they share their music with a global audience.”
Supertone
AI for Good said: “Jiseok, PJ, Chanyeon live with varying degrees of deafness and their dream to become an idol seemed unpromising until AI was adopted.” The group was formed April 2024 and they have since produced and performed music using a combination of AI tools and tactile technology such as vibration-based devices for choreography, The Korea Herald said.
Their debut track Glow was recognized as the first K-pop song to feature AI-generated vocal support, according to the news outlet.
The group has already gained recognition being named by Billboard as the “K-pop Rookie of the Month” in September, while Forbes included them on its “Asia’s 30 Under 30” in entertainment in May.
During a presentation at the UN AI summit, Haley Cha, CEO of Parastar Entertainment, said: “Our goal was to allow Big Ocean JJ to communicate naturally with global fans, without sacrificing their vocal character.”
“Supertone’s AI allowed us to separate tone and pronunciation, combining the strengths of both the artist and the guide vocal,” Cha was quoted by The Korea Herald as saying.
Supertone CEO Lee Kyo-gu said: “This collaboration shows how Supertone’s AI technology can support new artistic challenges in K-pop.”
“Our goal was to allow Big Ocean JJ to communicate naturally with global fans, without sacrificing their vocal character. Supertone’s AI allowed us to separate tone and pronunciation, combining the strengths of both the artist and the guide vocal.”
Haley Cha, Parastar Entertainment
“We’ll continue helping artists push past physical and technical limitations in their creative work,” Lee said, according to The Korea Herald.
Big Ocean is set to perform in Minneapolis, Las Vegas, Kansas City, New York, Washington, Raleigh and Atlanta through early August.
Supertone’s voice cloning process uses its NANSY (Neural Analysis & Synthesis) foundation model, which the company describes as “a unified neural framework designed to perform voice-related generative tasks.”
“Through its integrated structure, it maintains voice characteristics consistency across generation and allows for the control of four individual elements of voice, enabling the expression of all human voices in the world.”
Supertone has already established partnerships with Disney, Netflix, ZTX and ZEPETO, according to its website.
HYBE acquired Supertone in 2022, with plans to integrate its voice-creation tools into music. Less than a year later, HYBE-signed artist MIDNATT released a track called Masquerade, billed as the first-ever track to feature AI-generated vocals in a number of languages including Korean, English, Japanese, Chinese, Spanish, and Vietnamese.
In April 2024, the AI company released Supertone Shift, a tool that enables artists to alter and clone their voices in real time. The tool can be used, for instance, to turn a single artist’s voice into a chorus during a live performance.
HYBE also debuted a virtual pop group, SYNDI8 last year, using Supertone-generated vocals for all performances.
The K-pop giant demonstrated Supertone’s capabilities during an earnings call last year, revealing that then-CEO Jiwon Park’s investor presentation was actually delivered by an AI voice tool that replicated the executive’s speech patterns.
Music Business Worldwide
