Detecting interaural incoherence based on variations in the hemispheric balance.
* Presenting author
Human listener demonstrate a remarkable sensitivity in detecting evenslight incoherences within signals presented at the two ears. Theseinteraural decoherences can also be described as time and frequencydependent variations in both the interaural phase and level differencewhich are important cues for sound localization. Traditionally, theability to detect interaural incoherences is explained using delayline models where the incoherence results in a reduction in themaximum of the cross-correlation function. Conversely, by analyzingsingle-cell recordings from neurons in the inferior colliculus ofanesthetized guinea pigs, this study finds that interaural incoherencecannot be reliably detected when applying this mechanism. Instead, analternative detection mechanism based on a hemispheric balance modelis proposed, where stimulus power fluctuations result in covariantfluctuations in the two hemispheres so that the balance remainsunchanged, while fluctuations in interaural differences result inopposing fluctuations. The proposed mechanism thus considers thedisturbances in the hemispheric balance. By re-analyzing the recordedneuronal responses for variations in the hemispheric balance, it isshown that this approach considerably improves the performance ofdetecting interaural incoherence.