Where did that noise come from? Memory for sound locations is exceedingly eccentric both in front and in rear space.

Few studies have examined the stability of the representation of the position of sound sources in spatial working memory. The goal of this study was to verify whether the memory of sound position declines as maintenance time increases. In two experiments, we tested the influence of the delay between stimulus and response in a sound localization task. In Experiment 1, blindfolded participants listened to bursts of white noise originating from 16 loudspeakers equally spaced in a 360-degree circular space around the listener in such a way that the nose was aligned to the zero-degree coordinate. Their task was to indicate sounds' position using a digital pointer when prompted at varying delays: 0, 3, and 6 s after stimulus offset. In Experiment 2, the task was analogous to Exp. 1 with stimulus-response delays of 0 or 10 s. Results of the two experiments show that increasing stimulus-response delays up to 10 s do not impair sound localization. Participants systematically overestimated the eccentricity of the auditory stimulus by shifting their responses either toward the 90-degree coordinate, in alignment with the right ear, or toward the 270-degree coordinate, in alignment with the left ear. Such bias was analogous in the front and in the rear azimuthal space and was only marginally influenced by the delay conditions. We conclude that the representation of auditory space in working memory is stable, but directionally biased with systematic overestimation of eccentricity.