Purchase this article with an account.
Andrew Rider, Bruce Henning, Andrew Stockman; Light adaptation and the human temporal response revisited. Journal of Vision 2016;16(12):387. doi: https://doi.org/10.1167/16.12.387.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Light adaptation speeds up visual processing such that observers become relatively more sensitive to high-frequency flicker. Adaptation changes the temporal contrast sensitivity functions (TCSFs) from low-pass to band-pass: at low temporal frequencies contrast sensitivity remains approximately constant (Weber's Law), whereas at the highest frequencies it depends on flicker amplitude and therefore increases (high-frequency linearity). Several distinct models have been proposed to account for these complexities. Here, we re-evaluate them by modelling across existing TCSF data. Models that account for TCSF data are usually built up from low- and high-pass filters and typically differ in (i) the form and physical explanation of the filters, and (ii) which filters light adapt. Importantly, these models predict significant differences in how thresholds decline with frequency. One type of model predicts thresholds fall exponentially with frequency (consistent with the Ferry-Porter law); a second, based on solutions to diffusion equations, predicts that it falls with the square root of frequency; and a third, based on cascaded integrators, predicts that it falls with the power of frequency (with a slope equal to the number of integrators). Existing TCSF data vary significantly due to differing stimulus parameters and perhaps individual differences, but are most consistent with a Ferry-Porter-like asymptote. We demonstrate that a model based on cascaded integrators and appropriate feedback can produce an approximately exponential fall-off in threshold over a broad frequency range (15-60 Hz) tending toward a power law asymptote at frequencies that are too high to be visually insignificant (>100Hz). Our model contains two intensity-dependent terms, one controls the time constant of the low-pass filters and the other controls the overall gain to account for bleaching effects. The model also accounts for several other flicker perception phenomena including Weber's law at low frequencies, high-frequency linearity and the Ferry-Porter law in flicker fusion.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only