Rhea T. Eskew
Stephen G. Harkins (1948- ), Adam J. Reeves
Date of Award
Doctor of Philosophy
Department or Academic Unit
College of Arts and Sciences. Department of Psychology.
color vision, decrement, increment, masking, pedestal, S-cone
Eye--Accommodation and refraction, Eye--Adaptation
The short wavelength sensitive or S-cones are known to be both physiologically and psychophysically different than both the long and middle wavelength sensitive or Land M-cones. The L- and M-cones are genetically/evolutionarily similar to one another and both quite unique from the S-cones. The S-cones are also not found in the central retina where the L- and M-cones are most densely packed. This thesis uses a classic pedestal experiment to explore yet another unique feature of the S-cone visual pathways.
In a pedestal discrimination experiment, subjects must discriminate between a constant “pedestal” stimulus and another composed of a sum of the same pedestal plus a test. Pedestals can produce masking or facilitation. Masking is generally believed to result from contrast gain control mechanisms, which act to keep perceived contrast approximately constant under variable contrast ambient conditions. Here, pedestals of various chromaticities were used to mask and facilitate tests that are only detectable to the S-cones.
Psychophysical and physiological evidence indicates that S-cone increments and decrements are processed via qualitatively unique pathways, presumably S-ON and SOFF. For example, Study 1, employed a variety of pedestal chromaticities (S, L=M, L, M, -L=M, and Achromatic) with S-cone tests. High contrast S-cone increment (purplish) pedestals produced more masking than high contrast S-cone decrement (yellowish) pedestals. Similarly, an equal ratio of L- and M-cone increment (yellowish) pedestals produced far less masking than an equal ratio of L- and M-cone decrement (purplish) pedestals. These results indicate that the S-cone increment pathway may have a greater contrast gain control than the S-cone decrement pathway and that the two may be processed by distinct visual channels. None of the other pedestal chromaticities elicited any masking. If the equal ratio L- and M- decrement pedestals produced masking, but neither of the individual components that made up this pedestal did, the combination of cones must be nonlinear.
All pedestals and tests in Study 1 were displayed using a “rapid start” sawtooth waveform that rapidly increased (or decreased) in contrast and then linearly ramped back to the mean field. The incremental or decremental stimulus contrasts, coupled with “rapid-start” temporal waveforms, should differentially stimulate S-ON and S-OFF pathways. A second study explored whether the temporal sawtooth polarity is important in creating the asymmetry found in study I. The pedestal masking experiment was repeated with a “rapid-stop” temporal waveform that linearly ramped to a peak and then abruptly returned to the mean field. Even with high contrast pedestals, the sawtooth polarity had little effect, indicating that this temporal manipulation has little effect relative to the contrast manipulation.
A third study focused on another aspect of these models, having to do with the way cone signals are combined prior to the stage at which contrast gain control occurs. Various ratios of S-cone and L=M pedestals that were equated in cone contrast were used. If these cone signals were linearly combined with one another in the S-cone pathway, one would expect to find a symmetric pattern of masking as a function of the color angle of the pedestal. The pattern of masking for S+ is asymmetric indicating that the combination of S-cone and long-wavelength cones to the S+ pathway must be nonlinear. The patterns for S- tests are much closer to the linear prediction.
In conclusion, Study 1 showed a qualitative difference between S+ and S- that is consistent with physiological finding of two pathways with different contrast gain control (Solomon & Lennie, 2005). Study 2 showed that contrast polarity is sufficient to separate these pathways and further temporal manipulations provide little or no added benefit. Lastly, Study 3 proved that the cone combinations feeding into the S+ pathway are combined nonlinearly. This finding opposes the classic linear model that is often cited (Boynton, 1979) and adds to the plethora of unique features of the S-cone system.
Scott Huston Gabree
Gabree, Scott Huston, "Asymmetries in the s-cone increment and decrement pathways as revealed by pedestal masking" (2009). Psychology Dissertations. Paper 13. http://hdl.handle.net/2047/d20000143
Click button above to open, or right-click to save.
Additional FilesSupplemental Graph 1.avi (26724 kB)
Supplemental Graph 1