Supertone, the AI voice firm owned by Ok-pop large HYBE, has launched a brand new AI text-to-voice device that it says generates “high-quality, expressive audio content material.”
Dubbed Supertone Play, the brand new device is accessible in English, Korean, and Japanese. It will likely be expanded throughout the yr to incorporate service in Spanish and Chinese language, Supertone stated in a press release on Tuesday (February 18).
Supertone Play has a library of 150 voice characters designed for a number of content material creation makes use of, together with pure dialog, skilled narration for information and audiobooks, and “trend-driven meme voices.”
It additionally features a voice-cloning function that permits customers to create a voice character from their very own speech “in simply 10 seconds,” in line with a press launch. The voice character is customizable to imitate totally different emotional tones.
For these apprehensive the know-how may very well be abused to create superfakes, Supertone provides assurances: Its new device incorporates safeguards comparable to ‘Voice Captcha’ to make sure a creator’s voice – and never another person’s – is getting used to create customized characters, in addition to ‘inaudible watermarks’ to make Supertone-created vocals traceable.
The corporate says the device generates voices of enough high quality to make use of in video video games, commercials, movies, and podcasts. The device is accessible now on the Supertone web site, and is free to attempt till April 16.
“Play marks a serious step ahead in making AI voice know-how extra accessible and versatile for creators throughout the globe,” Supertone President Kyogu Lee stated.
“With its extremely expressive and pure speech synthesis, Play permits creators to carry their tales to life in new and dynamic methods. Whether or not for content material creation, storytelling, or immersive experiences, we’re excited to offer a device that enhances inventive expression by voice AI.”
“Play marks a serious step ahead in making AI voice know-how extra accessible and versatile for creators throughout the globe.”
Kyogu Lee, Supertone
Supertone stated that in its beta-testing part, the Play device was utilized in greater than 80 international locations and generated round 260,000 audio sentences. Of these, round 120,000, or 42%, have been downloaded, “a robust indicator that customers actively included the generated content material into their inventive initiatives.”
HYBE acquired Supertone in 2022, with an eye fixed to integrating its voice-creation instruments into music.
Lower than a yr later, HYBE-signed artist MIDNATTlaunched a observe known as Masquerade, which was billed because the first-ever observe to function AI-generated vocals in a number of languages. The observe featured lyrics sung in Korean, English, Japanese, Chinese language, Spanish, and Vietnamese.
Final yr, the AI firm launched Supertone Shift, a device that permits artists to change and clone their voices in actual time. The device can be utilized, for example, to show a single artist’s voice right into a refrain throughout a reside efficiency.
In an indication of Supertone’s capabilities, throughout an earnings name final yr, HYBE revealed that the investor presentation given earlier by then-CEOJiwon Park was truly delivered by an AI voice device that mimicked the manager’s vocals.Music Enterprise Worldwide
