Abstract
We used a dichoptic adapting and test paradigm to examine the functional organisation of adaptation processes in spatio-temporal vision. Observers were simultaneously adapted to two horizontal and two vertical flickering 1-d blue noise patterns arranged around a fixation point and presented to either the left or right eye (at random locations across runs) with a Wheatstone stereoscope. Adaptation periods were cycled with test intervals containing one vertical and one horizontal flickering noise pattern, whose locations (4AFC) were identified by observers. This arrangement produced 8 randomly interleaved combinations of vertical or horizontal adapt and test patterns, presented monoptically or dichopitcally. Threshold elevation was greatest when adapt and test patterns were of the same orientation and monoptic, with high temporal frequency (TF) adapting patterns producing greatest threshold elevation. Orthogonal effects of adaptation were approximately the same in magnitude under monoptic and dichoptic adapt-test pairings and in some cases produced cross-orientation facilitation — ie detection thresholds that were lower following adaptation to an orthogonal flickering pattern than with no adaptation. Our results are generally inconsistent with simple contrast gain control models involving two TF channels where the adaptability and sensitivity of the visual system are predicted to correlate, but consistent with the predictions made by an optimized two-temporal channel signal encoding and decoding strategy in which there also exists the possibility for whitening at high spatial and temporal frequencies. The observation that this shift is at least as great for dichopitc presentations suggests that the shift in TF tuning originates in cortex and may impinge on pre-cortical areas via feedback.
Supported by the Wellcome Trust and by the EPSRC