Abstract
Reading is tolerant to contrast reduction, but its rate decreases when character size becomes quite large or small (Legge et al., 1987). This study re-examines the relation between letter contrast and reading function for a broad range of luminance contrast. We measured the reading speed for single Japanese clauses comprising two-letter Kanji and two-letter Kana words in various sizes and luminance contrast levels. Reading rate (char/min) was calculated from the reading time for each clause. The largest clause was displayed in the first trial, and character size decreased by 0.1 log step until participants could not read any letter. The letter size was ranged from 3.10 to 123.86 arc min. The participants were 9 young and 20 older Japanese people with normal vision. Reading speed function of character size was plotted and three parameters—maximum reading speed (MRS), shape of the function, and the location of reading function (LRF)—were estimated for each contrast using Weibull fitting. While MRS and the shape of the functions remained unchanged, LRF systematically varied according to stimulus contrast, which agrees with the finding of Fujita et al. (2008). The relation of LRF and contrast were linear in log-log scale (LRF(log) = a + contrast(log) * b). Regression analysis indicated that participants' visual acuity had a statistically significant effect on its interception. Our results revealed similar but stronger tolerance of reading performance for a wide range of contrasts. The systematic dependence of LRF on luminance contrast reflects the relation of the critical spatial frequency for reading (Legge et al., 1985) and human CSF. With decreasing stimulus contrast, spatial frequency available to observers becomes lower but low frequency channels allow fluent reading.
Meeting abstract presented at VSS 2016