Purchase this article with an account.
Eleanor O'Keefe, Yeon Jin Kim, Edward Essock; Differential Properties of Narrowly-Tuned and Broad Temporal Channels. Journal of Vision 2013;13(9):272. doi: https://doi.org/10.1167/13.9.272.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The visual system contains multiple narrowly-tuned spatial frequency channels. In contrast, only two broad temporal channels are typically reported ("sustained" and "transient"). However, recent broadband masking experiments show multiple, narrow temporal channels in addition to the two broadly-tuned mechanisms (Kim and Essock, VSS, 2010). In the present work, we investigated the temporal tuning of flicker masking and cross-orientation masking (XOM) to better characterize these two types of temporal filtering. In the first experiment, test stimuli were high-speed or low-speed horizontal Gabor patches, which were masked by uniform-field flicker at varied temporal rates (2 to 20 Hz). The tuning of this flicker masking was measured for four test orientations (0°, 45°, 90° or 135°) to assess whether an anisotropy exists in masking by spatially unpatterned fields. In the second experiment XOM was examined using broadband masks consisting of 1/f broadband oriented noise that was either iso- or cross-oriented, and temporal tuning of masking was assessed. Flicker masking increased with temporal rate and showed, at best, a weak peak at ~10 Hz. Most importantly, flicker masking was anisotropic, with thresholds for horizontal test gratings higher than those of other test orientations (horizontal effect), even though there was no orientation information present in the mask. In the broadband masking conditions, temporal tuning shows two peaks: one around 10 Hz attributed to the high-speed ("transient") mechanism, and one attributed to the narrowly-tuned suppression pools reported previously. We suggest that the effects of many types of masking are pooled at the level of the cortical filters (perceptual channels) and that the anisotropy observed is due to differential weighting of the output of these normalization pools upon filters of different orientations (rather than anisotropic weighting of the inputs to these pools from filters of different orientations).
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only