Abstract
Purpose:
The standard Sloan letter set, which consists of 10 letters that are similarly identifiable when presented at a large size, has been incorporated into clinical charts that present the letters in additive white luminance noise. The present study determined if the addition of white luminance noise increases the variation in contrast threshold among individual Sloan letters.
Methods:
Contrast thresholds for letter identification were obtained from three visually normal subjects (ages 22, 25, 34 years). Letters from the Sloan set (C, D, H, K, N, O, R, S, V, Z) were constructed according to established guidelines. The letter size was equivalent to 1.5 log MAR and the letters were presented for an unlimited duration against a uniform adapting field or in the presence of additive white luminance noise. Contrast threshold for each letter was determined using an interleaved staircase procedure. To determine if the spatially broadband nature of the letters affects inter-letter thresholds, contrast threshold was measured for letters that were spatially bandpass filtered into 4 different frequency bands with peak frequencies of 1.9, 2.5, 3.8, and 7.5 cycles per letter.
Results:
For standard unfiltered Sloan letters presented against a uniform field, contrast threshold for individual letters differed by as much as 0.18 log units, consistent with previous reports. When measured in luminance noise, the maximum difference among the letters increased to 0.30 log units. A t-test indicated that the inter-letter contrast threshold differences were significantly greater when presented in noise (t = 3.44, p < 0.05). Bandpass filtering the letters further increased the differences in contrast threshold among the letters compared to the unfiltered letters. Additionally, noise increased the inter-letter threshold differences of the bandpass filtered letters, consistent with the effect observed for the unfiltered letters.
Conclusions:
The addition of luminance noise increases the differences in contrast threshold among individual Sloan letters. Bandpass filtering the letters to limit their object frequency content does not make the individual letter thresholds more similar. From these data, subsets consisting of at least 5 Sloan letters that differ by less than 0.15 log units can be constructed for the unfiltered and filtered letter sets presented in the presence and absence of luminance noise.
Keywords: 478 contrast sensitivity •
641 perception •
640 pattern vision