iHeartMedia Picks Synthetic Voice Developer Veritone to Translate Podcasts
Audio content giant iHeartMedia has picked synthetic speech startup Veritone to start translating the podcasts it hosts into new languages while retaining the original voices. Shows on the iHeart Podcast Network will begin appearing in multiple languages, starting with Spanish versions of the most popular English-language podcasts.
Veritone’s technology can analyze audio recordings and produce an algorithm that mimics the voice saying words and sentences not recorded by the speaker. Once the voice model is ready, Veritone has tools to imitate not just the sound of the voice but the pace, emotion, and other elements that make it realistic. The AI can translate scripts into different languages and make it seem like the original speaker is now fluent in a foreign tongue. iHeartMedia wants to apply that service to its podcasts, initially just in Spanish and only for its ‘‘marquee’’ shows, but eventually, for any language the AI supports. As part of the process, the platform ensures permission is given for the voice clone by the original speaker and opens new potential income sources for the podcast producers.
“With the hyper-growth of our podcast network over the past decade, we are always looking for what is new and next in the medium, too – and Veritone’s synthetic voice solution is a great example of that,” iHeartMedia Digital Audio Group CEO Conal Byrne said. “Being able to easily offer our podcasts in other languages, in the talents’ unique style, will be an awesome and innovative way to grow market share in the global marketplace.”
Voice Clone Boom
Veritone first became well known for its aiWare platform with enterprise services like audio and video transcription processing and analysis. The company has broadened its interests over the last few years into synthetic voice and related projects, integrating speech recognition technology from Speechmatics in 2020 and launching the Marvel.ai platform to streamline the creation and licensing of digital voices last year. The Veritone and iHeartMedia have been partners in several projects, but the new collaboration is a step beyond that earlier work. The voice clone tech will even be used in the advertisements in the podcast.
“We have worked with the iHeartMedia team for more than four years in various aspects of their business, and we are very excited to collaborate on expanding their podcast audience and unlocking new revenue streams for the broadcast industry leader,” Veritone president Ryan Steelberg said. “iHeartMedia has always been a pioneer of innovation, and now with the power of Veritone’s synthetic voice solution, iHeartMedia will not only be able to scale to new markets with localized language translations but retain the brand value of their top talent’s voice–which is fundamental in podcasting. We are also partnering to develop synthetic voices for advertising and engaging content while reducing time-to-market and production costs for radio, podcasting and the metaverse.”
The podcast translation concept is one of the many new ways synthetic voices are deployed in media, especially the new metaverse platforms. Plenty of startups, such as Resemble AI and Supertone are raising funding rounds and partnering with celebrities to market synthetic voices, and digital voice interface creator ReadSpeaker is augmenting SoundHound’s Houndify voice AI platform to sound more lifelike. The advances in synthetic speech technology mean even free, limited tools can produce impressive results like this fan-made trailer for Skryim voiced solely by AI. The tech giants aren’t ignoring the space either. Amazon’s voice synthesis tech has already turned Alexa into Samuel L. Jackson and given KFC Canada’s Alexa skill the voice of Colonel Sanders. Meanwhile, Microsoft’s Custom Neural Voice service has made AI versions of everyone from Bugs Bunny to Flo from Progressive Insurance, and Google is producing artificial voices for call centers.