Grasping lacks depth constancy in both virtual and real environments.

The use of virtual reality setups allows sophisticated manipulations of visual and haptic feedback important for the study of grasping movements. On this respect, the absence of feedback is shown to induce different systematic biases in visuomotor tasks, such as underestimation of the target distance and, above all, lack of the target depth constancy. The latter is revealed by a decrease in grip aperture with target distance increase, thus resembling biases in perceptual estimates of relative depths. However, whether these findings are mostly caused by the use of virtual reality rather than by an intrinsic property of the visuomotor system is still an open and debated question. We addressed this issue by comparing grasping movements performed in three experimental conditions: lights-on (L), glow-in-the-dark (G) and virtual (V). In condition L, a natural grasping towards the object was performed with the lights on; in condition G, feedback of the hand and object were rendered covering the distal phalanges of thumb and index finger and the object with glowing material; in condition V, the visual feedback of the hand and object was virtually provided and only defined by binocular cues. A spherical object was located at four different distances (420, 450, 480, 510 mm) and 16 subjects performed 40 trials for each condition. Results indicate a difference between conditions related to the duration of the movement but, strikingly, they show a clear and strong effect of object distance on the grip aperture in all the conditions equally. Specifically, grip aperture gradually decreased with increasing object distance, therefore revealing a clear lack of depth constancy. These findings demonstrate that systematic biases in grasping actions are not a prerogative of virtual environments and that the visuomotor system is subject to the same biases affecting perceptual estimates. Meeting abstract presented at VSS 2015.