In preliminary studies, we attempted to isolate motion mechanisms by measuring threshold contours for motion (direction identification) in L,M cone contrast coordinates. We discovered,in addition to the anticipated luminance motion mechanism (LUM), a spectrally-opponent motion mechanism (SPO). SPO responds to the difference of L and M yet appears distinct from the opponent L,M mechanism which signals hue (red-green, RG). We will thoroughly measure the properties of LUM, SPO and RG, especially quantifying differences between SPO and RG through their detection contours, the relative phases of their L,M,S cone inputs, and their motion adaptation properties. Initial studies suggest that LUM and SPO each extract an opponent-motion signal (e.g. response to difference of right and left moving components), and the net signals from the two motion mechanisms then combine linearly and independently of phase (indicating early motion analysis by separate mechanisms). We will study in detail this early-motion process and subsequent linear combination, and will examine whether motion adaptation and masking transfer across the LUM and SPO mechanisms. We also intend to continue the difficult task of isolating psychophysically the two postreceptoral hue mechanisms red-green (RG) and blue-yellow (BY) in order to quantify their properties: (1) The interaction of the three cone types (L,M,S) at threshold; (2) The effect of temporal frequency on the relative phase of the cone inputs; (3) Any ability to signal motion. Psychophysical isolation is difficult and we have developed procedures which take advantage of disparate cone phase shifts between mechanisms. Threshold contours will be reported in cone contrast coordinates to assess the relative cone weights on a physiologically plausible scale. We hope to resolve two controversial issues: what are the L,M weights for the yellow component of BY; what is the strength of the S input to RG. A significant S input to RG could permit RG and BY alone to explain discrimination of just-perceptible stimuli in the equiluminant plane (without 'higher-order' mechanisms).
Showing the most recent 10 out of 19 publications
