Detection of large interaural delays and its implication for models of binaural interaction.
Level 5 - mechanism / opinion, no new human data
Animal electrophysiology, comparative psychophysics, and computational modeling (bench/basic science).
PubMed 12083726 · doi:10.1007/s101620020006
What was done
Headphone-delivered broadband and narrowband Gaussian noise bursts (100 ms duration) with interaural time differences (ITDs) exceeding natural head-shadow limits were presented to owls and humans. Researchers assessed behavioral left-versus-right ear discrimination and recorded ITD-tuning curves from optic-tectum neurons in owls. A binaural cross-correlation model with short delay lines was constructed to simulate behavioral detection based on side peaks of ITD-tuning curves.
What was found
Owls behaviorally discriminated left-leading from right-leading sounds at ITDs up to five times their maximum natural delay. Electrophysiological recordings showed optic-tectum neurons were tuned strictly within natural ITD limits and did not respond to delays as large as those behaviorally resolved. The computational model using curved cross-correlation trajectories accounted for the large ITD detection and predicted sound localization reversals in both owls and humans. The abstract reports no sample sizes or statistical test values.
Why it matters
It demonstrates that sound localization at unnatural or extreme interaural delays does not require dedicated long neural delay lines, showing how side peaks in cross-correlation networks can account for spatial perception and localization illusions.
Limits
The abstract omits sample sizes for both human and owl cohorts, as well as exact quantitative metrics, variance, and p-values. Direct electrophysiological recording was limited to owls, meaning underlying neural mechanisms in humans were inferred rather than directly recorded.
Cited by
- supports The maximum interaural time difference for sound localization in humans is approximately half a millisecond.