Significance: Image simulation is a useful and efficient tool to explore the impact of spherical and astigmatic blur on visual acuity (VA) and image gradation. It could help to design new optical corrections more efficiently and rapidly.
Purpose: The purpose of this study was to compare the effects of simulated (convolution by an artificial eye) and real spherical and astigmatic defocus on VA and image gradation.
Methods: Experiments were performed under highly controlled conditions: dynamic correction of the subjects' aberrations at 1 Hz and application of an artificial pupil. In experiment 1, Landolt C VA was measured in various conditions of spherical and astigmatism defocus. The amounts of spherical or positive astigmatic defocus oriented at 45° that gives a Landolt C VA of 0.0, 0.2, and 0.5 logMAR were measured in experiment 2. In experiment 3, the subjects scored the quality of the perceived image (three high-contrast 0.4 logMAR letters) with a five-item continuous grading scale.
Results: Simulated blur was always more detrimental than optical blur. We measured a difference of 0.08 ± 0.03 and 0.11 ± 0.05 logMAR between both conditions, respectively, in presence of spherical and astigmatism defocus. An average ± standard deviation difference of 0.16 ± 0.06 D (i.e., spherical defocus) and 0.24 ± 0.15 D (i.e., astigmatism defocus) was observed between simulated and real optics blur to provide a given VA. The differences of image quality score between both conditions were, respectively, 15.13 ± 9.63 and 13.33 ± 4.83 for spherical and astigmatism defocus. Most of the differences were statistically significant.
Conclusions: We observed a difference of about 20 and 35% between simulated and real optics blur, respectively, in presence of spherical and astigmatism blur. However, the difference between both methods remains equal to or below the clinically significant difference.