Supertone, as long-time MBW readers will know, is the AI voice replication software program that Korean music big HYBE absolutely acquired in a $32 million deal in 2022.
On the finish of February, the AI capabilities of Supertone had been demonstrated through a voice clone of HYBE’s personal CEO, Jiwon Park, ‘talking’ on an organization earnings name.
Earlier at the moment (April 16), Supertone launched the beta model of a brand new software known as ‘Supertone Shift’.
The software is described as a “one-of-a-kind real-time voice changer” which HYBE claims will “empower creators with new prospects of expressions and content material creation”.
Image the scene: an artist singing reside on stage, through a number of completely different AI-assisted voices, all switched in actual time.
In Might final 12 months, HYBE showcased the chances of what it may do with Supertone’s expertise when it launched a brand new single known as Masquerade from an artist known as MIDNATT.
MIDNATT, the alter-ego of HYBE-signed artist Lee Hyun, used a prototype model of Supertone Shift to “seamlessly” transition between his personal vocals and the vocals of a feminine singer on the monitor Masquerade.
That very same monitor was touted by HYBE because the “first-ever multilingual monitor produced in Korean, English, Japanese, Chinese language, Spanish and Vietnamese”.
In response to HYBE, MIDNATT (aka Lee Hyun) sang the monitor within the six languages, and utilizing Supertone’s AI, “the pronunciation knowledge of native audio system was utilized to the monitor to additional refine the artist’s pronunciation and intonation”.
In response to Supertone and HYBE, the brand new software, Supertone Shift, permits customers to “instantly change to their desired voice” from a library of ten predefined voices in real-time, with their voice analyzed for conversion “inside a mere 10 seconds”.
Customers can then customise the voice by adjusting a mix ratio in addition to voice parameters comparable to pitch, pleasure, and reverb with “easy and intuitive sliders”.
In response to HYBE and Supertone, this stage of management “helps customers craft their distinctive type whereas sustaining their distinct sound, completely aligning with their artistic intent”.
You may see an indication of how the brand new software works within the demo video beneath. Supertone says within the video that it “may even make a customized voice for you” in case you can’t discover the precise voice you need from its presets.
HYBE and Supertone declare that Supertone Shift “represents a brand new frontier of real-time voice conversion expertise (RTVC) with latency all the way down to 47 milliseconds and hyper-realistic voice high quality, making it an optimum alternative for reside content material and performances”.
Immediately’s Supertone announcement says: “Whether or not VTubers, reside streamers, and podcasters present elevated audio experiences for listeners, or avid gamers and costume gamers immerse their followers in interactive voice chats, creators can seamlessly shift to their alter ‘identification’ on-line.”
Throughout the demo video above, Supertone confirms {that a} person “may even sing in [their chosen] voice on stage”.
The beta model of Supertone Shift is presently out there at no cost downloads as a desktop software till June 26, 2024 through its official web site, with its official launch slated for the latter half of this 12 months.
“By means of Supertone Shift beta, our aim is to gather suggestions from an array of creators to include a broader vary of voice samples and extra options to raised serve varied content material creation environments.”
Kyogu Lee, Supertone
“By means of Supertone Shift beta, our aim is to gather suggestions from an array of creators to include a broader vary of voice samples and extra options to raised serve varied content material creation environments,” mentioned Kyogu Lee, President of Supertone.
“We hope this innovation will elevate the worth of voice throughout the artistic course of, pushing the boundaries of creativity for creators.”
One other product launched just lately by Supertone arrived in November 2023 within the type of Supertone Clear, a real-time AI audio plugin for voice separation and noise discount.
Supertone says that the software has been used broadly by creators together with post-production engineers, video editors, streamers, musicians, and file producers in 198 nations.
The corporate reviews that it garnered over 27,000 month-to-month energetic customers inside simply six months of its launch in November, 2023.
Supertone first generated world media consideration in January 2021 with its so-called Singing Voice Synthesis (SVS) expertise.
The corporate used this tech to “resurrect” the voice of South Korean folks celebrity Kim Kwang-seok, with the next AI-generated voice debuted on Korean tv present Competitors of the Century: AI vs Human.Music Enterprise Worldwide