A theory is proposed that explains the substantial differences among the various measures of spatial acuity. According to this theory, the visual system can localize stimuli only to within regions that are several minutes of arc wide. Other measures of spatial acuity have limits that are one (two-line resolution) and two (localization) orders of magnitude finer because they allow the coarse positional labeling to be supplemented by detection of changes in contrast within limited bands of spatial frequency. It is argued here that the use of an intensive dimension (contrast), as a supplement to location, taps visual abilities that are not part of the position sense. The failure of traditional theories to reconcile the differences in the observers' ability to discriminate changes in contrast in the presence of the different acuity targets. Discrimination sensitivity in an array of contrast-sensitive filters varies with acuity targets and thus precludes a fixed transformation between contrast and position.
Support for the theory is obtained both from experimental data and from simulations. Positional labeling was measured by testing the ability of observers to correctly identify the location of a single line. Both resolution and localization thresholds were measured under a range of contrasts. In addition, sinewave masks at a variety of spatial frequencies and phases caused the changes in perceived offset required by the theory.
The model, made up of an array of spatial filters, each with a one octave bandpass, accounts for both the absolute sensitivity to positional offset and the variation of threshold with stimulus configuration. The parameters in a vector magnitude formulation of probability summation are derived from sinewave contrast sensitivity data. The detection model is extended to discrimination by the addition of a Weber response compression mechanism.
University of Rochester, March 1985.