Guest
Peder Jørgensen
Peder Jørgensen is Founder at Hance.ai, as introduced on the show.
Episodes (1)
Peder Jørgensen, founder of Hance.ai and co-founder of Soundly, explains how machine learning for audio enables noise removal and live stem separation.
Questions these episodes answer
How does Hance's machine learning run on limited hardware?
Jørgensen says Hance builds its machine learning to run directly on edge hardware, meaning the audio devices themselves, which are far more limited than cloud servers. That constraint is what lets Hance handle tasks like noise reduction and stem separation locally on a device rather than needing to send audio elsewhere to process it.
From: Pioneering Next-Generation Audio Enhancement with Hance.ai
What is stem separation and why is it useful for music?
Stem separation pulls individual elements, such as vocals or instruments, out of a finished music mix rather than working from separately recorded tracks. Jørgensen says Hance's version of it can run live, and calls hearing something like an isolated vocal from a well known mix a genuinely striking result of the technology.
From: Pioneering Next-Generation Audio Enhancement with Hance.ai
Why did Hance's dialogue noise-removal plugin become so popular?
Jørgensen says the plugin, called Extract Dialogue, took off quickly because it could run directly on a user's own device instead of requiring a slow, image-recognition-based approach to machine learning that had been common at the time it launched, in 2020.
From: Pioneering Next-Generation Audio Enhancement with Hance.ai
When was Hance founded and by whom?
Jørgensen co-founded Hance in 2020 together with Stian, a fellow Norwegian audio engineer whose company, Acon Digital, made highly regarded plugins used in Hollywood productions to alter recorded voices, bringing deep signal-processing expertise to the new venture.
From: Pioneering Next-Generation Audio Enhancement with Hance.ai
How are people using Hance's stem separation technology for karaoke?
Jørgensen says once Hance could isolate vocals live from a music mix, the team started holding karaoke parties in the studio, opening Spotify or Apple Music for the lyrics and singing along to any song with the vocal removed.
From: Pioneering Next-Generation Audio Enhancement with Hance.ai