HRTF diversity problem

HRTF databases - the measurements used to model how sound reaches ears from different directions - are typically collected from small, non-diverse populations. Mostly Western, mostly similar body types.

The result: spatial audio systems that work better for some ears than others, presented as a neutral technical standard. The default listener is someone specific.

Related: the AES has been working on more diverse HRTF databases but it's slow going. Some research into personalized HRTFs using ear geometry or machine learning to estimate individualized responses without full measurement sessions.

→ whose ears are the default · spatial communication beyond vision