Abstract
Purpose. Sensitivity to orientation differences is influenced by stimulus size, eccentricity of stimulus presentation and stimulus contrast. Mäkelä et al. (1993, Vision Research) studied the effects of size and eccentricity on orientation thresholds. They presented stimuli at a fixed high level of contrast and found that the threshold versus size functions obtained at each eccentricity could be collapsed onto a single function by dividing stimulus size at eccentricity E by F=1+E/E2, where E2 was about 1.95. Sally & Gurnsey (2001, ARVO) found that the recovered E2 was essentially unchanged when stimulus contrast was eliminated as a potential source of variability. Recently, Melmoth et al. (2000, IOVS) demonstrated that size and contrast make different contributions to form discrimination thresholds as a function of eccentricity, and proposed a method to recover these two eccentricity-dependent limitations. The purpose of the present study is to compare previous estimates of E2 for orientation discrimination with estimates obtained using the Melmoth et al. technique.
Method. Stimuli were vertical and tilted (1.5°) lines displayed on a high-resolution monitor. Contrast sensitivities were measured for identification of the tilted target as a function of size (0.75°–24°) across eccentricities (E = 0°, 2.5°, 5° and 10°).
Results. As stimulus size was increased, contrast sensitivities reached the same asymptotic level at each eccentricity. Scaling stimulus size with an E2 of 3.49° provided the best fit to the data.
Conclusions. In this experiment only size limited performance and there was no need to “double-scale” the data. It is possible, however, that contrast scaling may be required for a more challenging orientation discrimination task (e.g., a smaller orientation difference).
Supported by NSERC and FCAR grants to R. Gurnsey.