Head Mounted Display Study:
What’s Wrong With Your Head Mounted Display?
by Edinburgh Virtual Environment Laboratory, Department of Psychology University of Edinburgh

© 1993 Edinburgh Virtual Environment Laboratory, Department of Psychology, University. of Edinburgh ( john1@hms01.hms.uq.oz.au) and CyberEdge Journal

Originally published as a CyberEdge Journal Monograph
A supplement to issue #17 of CyberEdge Journal, September/October 1993
_______________________________________________________________

This is a summary of research accomplished over several months [Summer 1993] at the University of Edinburgh. A full report will be forthcoming, but due to the importance of this subject, we felt that our readers would want this information as soon as possible.

Abstract

The increasing use of Head Mounted Displays (HMDs) and the likelihood of their ever greater application causes concern regarding possible health and safety complications of their use. This study examines potential problems resulting from the use of current HMDs. Specifically, we find symptoms of binocular stress among HMD users.

The ultimate display

In 1965 Ivan Sutherland [1] presented a research report discussing “the ultimate display”. The prototype of this used two small CRT screens viewed through prisms and magnifying optics. It is worth noting that although this HMD did have the potential for stereoscopic presentation Sutherland [2] placed considerable emphasis on the “kinetic depth” cues the display could provide rather than disparity cues. This prototype, however, set the format for a host of HMDs that followed and most of the currently available/used models conform to a very similar design. Here-in lies a problem that arises out of a simplifying assumption that “if we place two-dimensional images on the observer’s retinas, we can create the illusion that he is seeing a three-dimensional object” (Sutherland [2]). This statement appears correct, but we would modify it by stating that “an illusion of a 3D object can be produced, but 3D space can not be rendered with integrity from dual 2-D images”. The implications of this re-statement arise when we replace the static stereogram, which provides an illusion of surfaces at defined depth increments, with a representation of a 3D world with objects at a full range of disparity increments, which the observer then attempts to selectively sample.

Research findings on HMDs

There have been a number of observations that stereoscopic display devices may produce some discomfort: there have been problems observed after using military night-vision goggles [3], anecdotal reports of stress following viewing stereo-screens [4], and a cluster of observations that HMD use produces some mixture of nausea/headache/tried eyes. Unfortunately there has been very little objective empirical research on the visual and ocular effects of wearing current HMD devices. Our own research [5] initially used the VPL LX EyePhone. We took 20 young adults, free from ocular disorders, sat them on a “tethered” bicycle and presented them with the task of cycling around a virtual road network that spanned a rural setting which included buildings, bridges and road signs. We chose this task because it requires that the subject samples both far visual space, to see where they are heading, and near space to negotiate tight bends or view signs. All the subjects were fully screened by our research optometrist prior to commencement of their session and after 10 minutes of low-intensity cycling. The results were alarming: measures of distance vision, binocular fusion and convergence displayed clear signs of binocular stress in a significant number of the subjects. Over half of the subjects also reported symptoms of such stress, such as blurred vision and for one subject it took 40 minutes for normal vision to occur. The initial response from some of the VR community was to question the wisdom of using the VPL LX and to blame the problems on the its use of Fresnel lenses. We have recently confirmed these results, however, with a similar HMD that uses LEEP optics, with over 50% of our sample showing associated problems. Our current research is also highlighting related problems with the use of stereo-desktop presentations for quite short viewing periods, which extends earlier observations of ocular problems following viewing durations of 1 hour [6].

The fundamental problem with HMDs

So what’s going wrong ? Well, the bad news is that we believe that there is a fundamental flaw in the design of current HMDs. Although dual/split screen presentations provide disparity cues that specify surfaces in depth, the image can only be seen clearly by accommodating (focusing) to the depth of the virtual screen image. A number of HMDs have retained the original focal setting of Sutherland’s prototype and set the screen image to around 50cm. An initial effect of this is that accommodation promotes a similar degree of convergence (rotation of the eyes inward), which coupled with “proximal” convergence results in a prolonged vergence effort, which has been shown to be related to visual fatigue [7]. This could be relieved by moving the virtual image depth out towards infinity, but this in turn amplifies a second problem: both accommodation and disparity give powerful cues for depth, and vergence eye movements are driven by both sources of information. They are physiologically linked. It has been observed previously that “the user must learn to decouple accommodation and convergence” (Robinett & Rolland [8] ) but the implications of this suggestion have not been given enough weight. Such decoupling can, at best, only be a transient state and the visuo-motor system will constantly struggle to respond to the differing inputs of blur and disparity. Eventually a system breakdown is inevitable. In this respect, moving the virtual image toward infinity does not assist, because it merely increases the difference between blur-specified and disparity-specified depth when the user is studying close objects.

What is it HMDs really need ?

Everyone accepts that increased screen resolution is a requirement for future HMDs, but equally we would suggest that a minimum requirement for the reduction of serious visual stress in stereoscopic presentations is variable focal depth. This is not a minor issue, the technology to support this is very patchy. There are two approaches to the problem. A display system could “paint” the image so that rays specifying a given object require accommodation equivalent to the virtual depth of the object. In principle this is possible with a very fast oscillating lens system, but in practice is not presently feasible. Or, the full image depth could change as a result of the user's point of regard (the depth of their fixation). This would mean servo-ing the HMD optics to measurement of the user's eyes. Again this sounds feasible, but stumbles on the issue of practicality. Accurate eye tracking systems (Purkinje tracking, magnetic coil) are very cumbersome or uncomfortable, whereas small unobtrusive eye-tracking systems (limbus tracking, EOG) have poor stability. In addition, all such systems have inevitable transmission delays, and the speed of eye movements make the problems of lag in tracking the (slow moving) head very minor in comparison. One partial solution may be the sensible use of an HMD with manually adjustable focus. There is at least one HMD with this facility currently in production. Our proposal is that the user should establish the focal settings that would result in near - mid - far virtual image depths, allowing for their own visual prescription. The user should set the HMD image appropriately for the work to be attempted (near setting for proximal tasks, far setting for distance viewing). The image generation software should ideally “cull” disparities that specify objects outside of this working range. We believe this should then significantly reduce the visual stress induced by HMD use. We have not, however, been able to acquire a suitable HMD to confirm this proposal and as such it should be taken as tentative suggestion, support for which has yet to be provided. In the meantime we would emphasize that there is strong evidence that HMD use does induce significant binocular visual stress.

One final solution is to move to a monoscopic system. LEEP’s Cyberface3 HMD would seem to potentially avoid most of the problems discussed, although again we have not evaluated this display. The issue is whether a stereo system is important for your application. In natural settings, disparity cues are clearly used for depth judgments in peri-corporal space. For an example, try picking up your toothbrush and squeezing on toothpaste, with one eye closed. Although kinetic depth cues can make observers in a cinema flinch as an object looms on the screen, they cannot make the object appear to project forward from the screen. A monoscopic display system will always have difficulty in presenting close working images for manual interaction tasks. The requirements of stereo and mono displays are application specific, but stereo systems clearly have an important role to play.

A Final Warning

Those using HMDs should carefully consider the duration and repetition of their use and restricting their use by those with visual disorders or immature visual status. We can merely report our objective findings. The emphasis has to be on HMD manufacturers to confirm that their product does not produce visual stress. Until such data is available, tread carefully when you place an HMD on others. Disclaimer: The recommendations and data presented in this report are the result of controlled research studies under the supervision of a qualified optometrist. We cannot accept responsibility for the actions of others on the basis of this data. All recommendations made within this report are qualified to the extent that individuals may respond differently to display devices and no single procedure can be considered as constituting safe usage.

Edinburgh Virtual Environment research group includes: John Wann, Simon Rushton, Mark Mon-Williams, Rycharde Hawkes & Martin Smyth. The laboratory is funded by research grants from the UK MRC and SERC to Dr. John Wann & Prof. David Lee for research on principles of perception and action in natural and virtual environments.

References:

1. Sutherland I.E. (1965) The Ultimate Display. Proc. IFIPS
2. 506-508 2. Sutherland I.E. (1968) A head-mounted three dimensional display. Fall Joint Computer Conference, 33, 757-764
3. Sheehy, J.B. 7 Wilkinson, M. (1989) Depth perception after prolonged usage of night vision goggles. Aviation Space and Environmental Medicine. June, 573-579
4. Patterson R. & Martin W.L. (1992). Human stereopsis. Human Factors 34(6), 669-692.
5. Mon-Williams M., Wann, J.P & Rushton S. (in press) Binocular vision in a virtual world: Visual deficits following the wearing of a head-mounted display. Opthal. & Physiol. Optics.
6. Nishijima, T (1993) 3D images bad for eyes. The Daily Yomiuri, July 7th, p.9 (Newspaper report).
7. Tyrrell, R.A. & Liebowitz, H.W. (1990) The relation of vergence effort to reports of visual fatigue following prolonged near work. Human Factors 32(3), 341-357.
8. Robinett W. & Rolland J.P. (1992) A computational model of the stereoscopic optics of a head-mounted display. Presence 1 (1), 45-62.

Questions? More Info? Email the Webmaster

407 M. L. King Jr. Way, Oakland, CA 94607 USA
+1 510 419-0800

Our privacy policy may be read here.