Abstract
Conventional wisdom is that accommodation in humans exhibits significant errors. When the stimulus is far, the eye is thought to focus too near (“lead of accommodation”). When the stimulus is near, it focuses too far (“lag”). These errors are as large as 1 diopter, which should produce noticeably blurred imagery. But viewers typically do not experience the blur expected from such leads and lags. We re-examined this phenomenon by measuring accommodation objectively and subjectively. Objective measurements are based on measurements of light reflected off the retina. Subjective measurements are based on the viewer performing a visual task; they are more valid because they involve the whole visual process.
We used a custom varifocal display apparatus to present accommodative stimuli to six young adults. On each trial, subjects fixated and focused on a Maltese cross at a distance of 0, 2, 4 or 6D. A wavefront sensor measured accommodation (and pupil size) objectively. During the 3-sec presentation of the cross, we also measured acuity with flashed tumbling E’s presented at different distances. Subjects indicated the E’s orientation at the end of each presentation. Percent correct was determined for each of eight distances of the E relative to the accommodative stimulus. Each subject experienced 1600 trials (4 accommodative stimulus distances, 8 E distances, and 50 repetitions).
The objective measurements from the wavefront sensor revealed typical lags and leads. The subjective measurements revealed much smaller lags and leads. Thus, by subjective measurement, the eye appeared to be focused quite accurately at the distance of the fixation stimulus.
We conclude that errors of accommodation have been overestimated. The eye is better focused at the stimulus distance than previously thought. Our observations have important implications for the understanding of accommodation, the development of myopia, and next-generation head-mounted displays.