I did my thesis work on human color vision and visual sensitivity, working at the Ohio State University with Carl Ingling in the Department of Biophsyics. While the experiments did not work out as planned this led me to my abiding interest in how the visual system worked and led to my postdoc. My postdoctoral studies were at the University of Chicago with Joel Pokorny and Vivianne Smith in the Department of Ophthalmology. There I got to work on both color discrimination, and a topic that kept cropping up in my career, nonlinearities in visual processing. This was specifically our studies of the Abney Effect. I also had the opportunity to interact with residents and take specialized courses that ranged from clinical topics to basic science.
When I moved to the University of Pittsburgh I needed a new project and realized that it was possible to make physiological measurements using psychophysics. I started writing grants and set up a lab in the back of the Eye Bank that measured the change in the spectral sensitivity curve of the long and middle wavelength cones that occur when they are bleached. These changes arise from the change in absorption spectrum with a change in the optical density of a pigment (Beer's Law), an effect also known as self-screening. In the cones this provides the experimenter a way to psychophysically measure cone photopigment bleaching and regeneration, just by making color matches. We got this idea funded by NIH on the third try, and began a period of very productive studies, first to validate the technique, and then to apply it to retinal diseases, including RP, CSR and diabetes, as well as normal aging and AMD.
Our intererest in cone responses led to a goal of measuring the cones using the ERG, with the basic idea being that the early receptor potential should be the fastest response in the retina, and thus high speed flicker could be used. This didn't quite work out because we found that we could measure the flicker ERG out to almost 200 Hz with suitably sensitive signal processing, But we did get interested in the fact that the ERG response was highly non-linear, and we could use these nonlinearities to tease apart stages of visual processing (this uses a sandwich model of nonlinear processing). This interest in nonlinearities in visual processing led us, working with Adam Reeves, and Rhea Eskew, to investigate some psychophysical correlates of the fast retinal changes as well as to compare the filtering seen in the ERG to that measured using psychophysics, again with the goal of separating out stages of retinal and later processing. The work was quite useful, and led to our idea that a lot of retinal processing, which is highly non-linear, is working to linearize the retinal output.