Backus, B.T. & Banks, M.S. (1999) Estimator reliability and distance scaling in stereoscopic slant perception. Perception 28: 217-242
Abstract
When a horizontal or vertical magnifier is placed before one eye, a frontoparallel surface
appears slanted. It appears slanted away from the eye with horizontal magnification (geometric
effect) and toward the eye with vertical magnification (induced effect). According to current
theory, the apparent slant in the geometric and induced effects should increase with viewing
distance. The geometric effect does scale with distance, but there are conflicting reports as to
whether the induced effect does. Ogle (1938 Archives of Ophthalmology 20 604 ^ 623) reported that
settings in slant-nulling tasks increase systematically with viewing distance, but Gillam et al (1988
Perception & Psychophysics 44 473 ^ 483) and Rogers et al (1995 Perception 24 Supplement, 33)
reported that settings in slant-estimation tasks do not. We re-examined this apparent contradic-
tion. First, we conducted two experiments whose results are consistent with the literature and
thus replicate the apparent contradiction. Next, we analyzed the signals available for stereoscopic
slant perception and developed a general model of perceived slant. The model is based on the
assumption that the visual system knows the reliability of various slant-estimation methods
for the viewing situation under consideration. The model's behavior explains the contradiction in
the literature. The model also predicts that, by manipulating eye position, apparent slant can be
made to increase with distance for vertical, but not for horizontal, magnification. This prediction
was confirmed experimentally.
