Many measures for the ``visual effectiveness'' of a virtual environment are possible in which the ability of the visual cues to influence spatial perception is evaluated. One can, for instance, measure the extent to which the perception of the gravitational vertical is influenced by conflicting visual vertical cues. For an explanation of why I consider these to be presence measures, see the discussion of the ``presence hypothesis'' in Section 3.3.3.
Two examples involving the perception of gravity are outlined below.
Hatada et al. [36] examined the ability of a display
to induce a change in the perceived vertical as a function of horizontal FOV
and scene content. They report very small tilts for FOV's less than
20-30
, and a saturation at about a 5 disturbance to the
perceived vertical when the display FOV reaches 80-90.
Nemire et al. [68] reported an experiment to measure ``simulation fidelity'' in terms of the ability of a VE to induce a change in perception of gravity-referenced eye level (GREL). A pitched optical array can bias a subject's estimate of eye-level. The authors report that a physical array biases GREL more than an identical virtual array. However, the addition of longitudinal (into the distance) lines to the virtual array removed the performance difference.
The research described in Chapter 4 deliberately factored gravity out by restricting motion cues to the horizontal plane. The reason is that gravity is a very strong cue: it is difficult for visual cues to overwhelm it, particularly when the visual cues are provided by a poor virtual environment. A more sensitive measure is possible when the influence of gravity is controlled for.