Virtual environments are used in a variety of simulation and training applications. The user’s perception of these environments inspires multiple research questions. One of the most important is: “How do we perceive distance in a virtual space?” The authors conduct an investigation into the accuracy of metric judgments in virtual space. They support, through three sets of experiments, the hypothesis that three-dimensional (3D) visualization systems where users are expected to discover relations based on Euclidean distances or shapes are ineffective.

The first experiment estimates the size of errors made when judging different 3D shapes in normal lighting conditions, and studies the relationship between the error size and the distance from the viewer to the object. A significant error in the perception of a 3D shape, with no relation to the viewer’s distance to the object, is confirmed.

The second experiment has a similar goal as the first, however the range of shapes is extended to include thinner and more elongated objects. Additionally, the shape errors in the depth direction are compared to the errors made in the front-parallel direction (namely, left/right to the viewer). The finding of this experiment is that judgment of metric shape is more error prone in three dimensions than in two dimensions, and that errors seem to increase as the shapes become more elongated in depth.

A third experiment tries to reduce the errors in metric shape perception by employing an identical comparison block method. The experiment is based on the assumption that a comparison block would allow users to scale depth in units of the block, and thus would improve perception. The authors find that using a comparison block method does not improve the metric shape judgment.

The overall outcome of the experiments emphasizes a large error in the perception of metric relationships between the objects in a virtual setting. The paper presents a realistic set of experiments, and associated analysis, in support of the idea that the human visual system cannot make accurate metric judgments in virtual space. Given that the display system typically degrades user perception (for example, visual acuity) within a 3D simulation environment, it is important to quantify these parameters to establish their relevance.