Going beyond Alex White's first paper on the topic
Discussing effect of light on latency
I think you're probably right that the Hess effect comprises at least two separate effects of luminance and contrast on latency. Alex White summarizes some of the data from Hess and Pulfrich in the email below, which seem to indicate a role for overall adaptation level.
I know of 3 effects, that often work together in natural conditions, that may underlie this:
1. I think the rods themselves are slower than the cones, such that even with equivalent contrast (if there is any way to equate contrast in the cone regime and the rod regime) the system would have higher latency in the rod regime. 2. even within the photopic range, the system gets faster with increasing luminance. Kelly (1961) showed that at higher mean luminance, the impulse response function inferred by flicker sensitivity had a faster response. 3. as shown by Stromeyer & Martini (2003), increased contrast itself also speeds up the impulse response function.
On 03/06/2008, at 12:08 PM, Alex White wrote:
As far as I can tell, visual latencies as measured by the Hess and Pulfrich effects, and by RTs, depend on both contast and luminance. The Hess effect works for both white targets on black background (Hess, 1904) and vice versa (Wilson & Anstis, 1969). However, latency differences for white targets on black backgrounds are bigger than for black targets on white backgrounds, according to Prestrude & Baker (1971) who used an experiment that was sort of a cross between the Hess and Pulfrich effects. This was also found by Kitaoka and Ashida (2007) with a Hess effect-like display.
From reaction time measurements, Plainis and Murray (2000) made a
model that includes contrast, luminance, and spatial frequency all as independent determinants of latency.
The level of adaptation of the retina seems to play a big role in determining latencies and could explain the relationship between contrast and luminance. Prestrude & Baker (1971) concluded that "adaptation is the major factor determining latency differences". In another experiment they bleached one eye by exposure to a bright light and then had subjects dichoptically view their stimuli and found that the bleached eye (that was adapted to higher luminance) had shorter latencies.
In other words, for one eye, the latency probably depends most on the space-average luminance level for one stimulus, which affects the adaptation state. Higher space-average luminance causes shorter latencies. I guess then for something like the Hess effect where 2 objects of different luminance are both presented to each retina, the space-average luminance for some sub-region of the retina will determine it's latency, but it also depends on the adaptation state that retina was in just previously.
Ok, thats my understanding of this issue that is more complicated than I first thought. Hope that helps.(Alex W)