Returning the Voice of a Singer with ALS
Patrick Darling is a 32-year old musician, composer, and singer. He lives with advanced ALS, which has taken many things from him including his ability to move, to speak and to sing. When speech-language pathologist Richard Cave set out to rebuild Patrick's singing voice using ElevenLabs' AI-driven voice cloning technology, the project depended on many essential factors, including clean source audio.
The problem was that Patrick's singing voice only existed inside fully mixed recordings like band tracks and phone videos, all with ambient noise bleeding throughout. No stems. No isolated vocals. And without clean source audio, even the best voice cloning technology will not produce good results.
Richard was introduced to AudioShake through Bridging Voice, who help people with ALS maintain their ability to meaningfully communicate. With AudioShake’s world-class stem separation, Richard was able to isolate Patrick's voice from those band tracks and phone recordings, accessing clean vocal stems from source material that would otherwise have been unusable. Using cutting-edge ElevenLabs technology, Richard used those stems to rebuild a singing voice that was so accurate it matched Patrick’s natural singing voice that he once had.
Using his ElevenLabs voice, Patrick sang an original song "Ghost of a Man I Never Met,” with his band playing live. All in front of nearly 1,000 people on stage at the ElevenLabs Summit in London.
"For the first time in years, I could hear myself sing again. That changes everything." – Patrick Darling
Patrick is planning an album and a tour. AudioShake continues to work with Richard Cave as he expands ElevenLabs AI-assisted voice restoration globally, building singing voices in 70+ languages like Hindi and Spanish for patients who have historical recordings of their natural singing voice.
Making sound usable when it matters most—enabling talent to shine including those living with speech loss - exactly what AudioShake is built for.
