Test of a model of auditory object formation using intensity and interaural time difference discrimination.

A model of auditory object formation and an experimental evaluation of the model are described. Specifically, predictions for intensity discrimination and interaural time difference discrimination for the central component of a three-component harmonic complex are evaluated empirically. The onset time of the central, target component is varied relative to the onset times of the remaining, interferer components in order to vary the degree of fusion (versus perceptual segregation) of the target and the interferers. The model, which is based on the idea of attenuation of the components in the nonattended auditory image (in the case of segregated images), predicts lower sensitivity to information at the target component for the fused versus segregated target, and equal sensitivities for completely segregated targets and targets presented in isolation. Results are presented for four subjects with component frequencies of 400, 600, and 800 Hz and with onset time differences of 0 or 250 ms. The target duration was always 100 ms and offset times were the same for all components. The subjective results were as expected, with synchronous onsets yielding one sound object, and asynchrony of the central component yielding two sound objects. Also, the empirical results on interference in the synchronous case were in qualitative agreement with the above predictions. However, significantly more interference was found than was predicted for both synchronous and asynchronous conditions. In fact, the amount of interference found contradicts the simple attenuation model of object formation.