Overview
- Macquarie University scientists report in eLife that listeners improved speech recognition over a single 45-minute session by learning the acoustic signature of a space.
- Tests in an anechoic chamber used real-room recordings from an underground car park, a lecture theatre and an open-plan office with background noise.
- Learning and intelligibility peaked around 400 milliseconds of reverberation, whereas very high echo or no echo at all hindered performance.
- Briefly disrupting dorsolateral prefrontal cortex activity with magnetic stimulation reduced adaptation, implicating specific neural circuitry.
- The findings suggest preserving some reverberation could aid hearing in public spaces and audio devices, with follow-up studies planned for neurodivergent listeners and people with hearing loss.