Instant Genius Inside the project helping the hard of hearing listen to music
May 3, 2026
Trevor Cox, acoustic engineer and professor at the University of Salford leading the Cadenza Project to improve music for people with hearing loss. He discusses why hearing aids struggle with music, the differences between music and speech for processing, challenge-style signal‑processing tasks and machine learning approaches to rebalance instruments and boost lyric intelligibility. Plans for consumer device improvements are also explored.
AI Snips
Chapters
Transcript
Episode notes
How Aging Erodes Instrument Clarity And Dynamics
- Age- and noise-related hearing loss commonly reduces high-frequency sensitivity and pitch perception.
- Trevor Cox notes this leads to trouble picking out instruments and loudness recruitment where loud sounds become distorted and unpleasant.
Hearing Aids Focus On Speech Not Music
- Most hearing aids and consumer devices are optimised for speech, not music, so music programmes are inconsistent.
- Trevor Cox says you cannot fully restore original hearing, so the goal is recreating the emotional 'oomph' of music.
Music Diversity Makes Processing Harder Than Speech
- Music is far more diverse than speech in frequency, dynamics and purpose, so processing goals vary by listener and context.
- Trevor Cox highlights differences from string quartets to metal to exercise playlists, each needing different treatments.