Holly Herndon is a 41-year old Sound Artist based in Berlin, Germany. She is partnering with music company Never Before Heard Sounds to produce a “digital twin” of her own voice, titled Holly+, that creators across the world can manipulate to create their own custom vocal parts and instruments. Similarly to our previously reported Jay-Z deepfake story, Holly has entered a multitude of different sounds in the Holly+ digital library for others to use to create their own custom works.
Holly+ uses “Machine learning technology to reproduce audio in a computer generated approximation of Holly Herndon’s natural singing voice.” Users can upload up to 5 minutes of polyphonic audio into Holly+ and that audio will then be recreated with Holly’s distinctive digital voice. Holly+ is a Decentralized Autonomous Organization (DAO) in which all users will get to own the IP for their creation, but all funds earned will be directed back into the organization to further the development of the service and will be shared amongst all artists on the site.
Holly claims that “Vocal deepfakes are here to stay.” To learn more and make your own creation using Holly’s voice, check out holly.plus.