Abstract
When alternated at low temporal frequencies pairs of coloured lights usually appear to vary in hue at their alternation frequency. As the frequency is increased, the hue variation fades to a steady mixed-hue appearance at frequencies well below the flicker fusion frequency, but a luminance variation of the mixed-hue remains (unless the alternating lights are luminance-equated). Traditionally, this loss of hue variation at relatively low frequencies has been assumed to reflect chromatic mechanisms' having greater temporal integration than the luminance mechanism. For several years, our group has been accumulating evidence for an alternative model that postulates that the hue variation is partly limited by the rate at which chromatic mechanisms can signal changes in hue (i.e., they are a limited by a "slew" rate). Our approach has been to vary the temporal waveform of flicker in ways that give rise to predictable changes in either the mean ("DC") hue appearance or the time-varying ("AC") hue variation that should result from a slew limit. These waveforms include sawtooth stimuli that vary in their on- and off-slopes from rapid-on to triangular to rapid-off, square-wave stimuli that vary in duty-cycle, combinations of 1st and 2nd harmonic sinusoidal flickering lights that vary in phase, and other more complex waveforms. The objective psychophysical tasks have included detection and matching of the DC and/or AC components. We have compared these psychophysical data with simulations. We find that a model in which the slew rate limit follows an expansive nonlinearly that accelerates input signals above and below the mean can account for many of the phenomena we have encountered in terms of changes to both the DC and the AC. Moreover, slew-rate models in which different unipolar or bipolar hue mechanisms have different slew rates may provide the basis for explanations of pattern induced flicker colours.
Meeting abstract presented at VSS 2016