Purchase this article with an account.
Andrew Glennerster, Stuart J. Gilson, Lili Tcheang; The representation of visual space in an expanding room. Journal of Vision 2005;5(8):197. doi: 10.1167/5.8.197.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We have found that when a virtual room expands around an observer as they walk, observers fail to notice any change despite veridical binocular and motion parallax cues (VSS 2003). This leads to gross errors in size constancy, particularly at far viewing distances. Here, we investigate whether this pattern of results imply a distorted representation of viewing distance in the expanding room.
Wearing a head mounted display in an immersive virtual reality system, subjects carried out the following forced-choice tasks in both an expanding and a static room. Each task was carried out with the subject on the left of the room (when the expanding room was small) and on the right (when it was 4 times larger). Subjects (i) compared the size of an object visible on the left side of the room with one visible on the right; (ii) judged the depth of a triangular prism relative to its half-width; and (iii) judged the distance of a rod as greater or less than half (or twice) the distance to a comparison rod. Only task (i) required the observer to compare two objects viewed when the room was a different size.
Estimates of viewing distance derived from the size comparison task were close to veridical in a static room but in the expanding room were consistent with either a compression or an expansion of visual space depending on whether the comparison object was shown when the room was small or large. For the depth-to-width judgement, subjects' biases were consistent with a compression of visual space in both static and expanding rooms. For the half- and double-distance task, judgements were close to veridical in both the static and expanding rooms. Conclusion: Performance in these tasks does not rely on a single internal 3D representation (e.g. Gogel, W.C., 1990). Instead, task-specific strategies are used.
This PDF is available to Subscribers Only