Posted with permission of the American Academy of Optometry:
"Depth-Discrimination Thresholds for Targets with Equal Retinal Illuminance Oscillating in a Frontal Plane",
American Journal of Optometry and Archives of the American Academy of Optometry,
vol. 43, issue 5 of May 1966, pages 283 - 298.
Copyright (c) The American Academy of Optometry
by John Michael Williams, 2000. Placed in public domain.
Stereoscopic acuity was studied in the Pulfrich apparatus without the
different illumination of the two eyes which causes the Pulfrich illusion.
Thus, the observer reported back-and-forth motion, with no apparent
ellipse--no near limb for target motion one way, nor a far limb
for target motion the other way.
Data for equidistance settings at several target speeds and levels of illumination are reviewed. The localization error (constant error or observer bias) is reported here, as is the stereoacuity (variable error) in the settings. The stimulus conditions and results were the same as reported in Lit (1964).
The discussion here emphasizes that the observer-specific ("personal equation") localization error probably always should be estimated and reported when studying the Pulfrich illusion at low illumination or high target speed.
The geometrical estimation of visual latency by the Pulfrich Effect always, if possible, should be based both on the apparent near and far limbs of the Pulfrich ellipse, and should be corrected for personal equation. However, in the data available at the time of this report, a small residual difference was found in the near-estimated latency versus the far-estimated latency, even after all known corrections had been applied.
The Pulfrich Effect, SIU-C. Last updated 2000-07-23