Can the amplitude difference spectrum peak frequency explain the foveal crowding effect?

Hess et al. (Vision Res. 40 (2000) 365) found that the peak frequency of the amplitude difference spectrum (ADS) of a Landolt C/flanking bar configuration was higher than that of an isolated C. They believed that this mismatch in the dominant frequency components could account for the foveal crowding effect. The following empirical and theoretical studies were conducted to evaluate this new explanation. First, the foveal crowding effect between a Landolt C and four flanking bars was measured under same- and mixed-polarity conditions. Significant crowding was found under the mixed-polarity condition. This result is different from the data that Hess et al. used to support their explanation of the foveal crowding effect. Second, analytical expressions for the Fourier transforms of C/flanking bar configurations were derived. ADS for various ring/bar separations and contrast polarities were calculated using these expressions. This analysis showed several discrepancies between ADS peak frequency predictions and empirical data. Therefore, the ADS peak frequency of the stimulus configuration does not provide an adequate explanation for the foveal crowding effect.