Lenny Lipton, Moderator, StereoGraphics Corporation

The panel discussion subjects went beyond its title and ventured into so many areas that what was discussed is difficult to summarize succinctly. There was a great deal of audience participation and the audience spoke nearly as much as the panel members. A lot of passion was expressed, which is not always evident during the presentation of the papers. I asked my fellow panel members to contribute their thoughts on the discussion, and three of them, Jeff Ferguson, Dan Sandin, and Mark Bolas set me their impressions. (Dave Cook was also a panel member.) In addition, Neil Dodgson was kind enough to contribute. (I take responsibility for editing the contributions.) As the reader will see, in some cases the panel participant went beyond the purview of reportage and provided comments or conclusions. The reader may get the impression that we held a number of different panel sessions. One thing to keep in mind, which may be have been forgotten, is that the production and projection of stereoscopic movies is ongoing and a commercial success, as every major theme park in the world has, I have been lead to believe, at least one stereoscopic cinema.

Jeff Fergasen, Ilixco Inc.

The panel seemed to be a success; the large audience participated and was genuinely interested in factors that will bring commercial acceptance and success to the stereo 3D business. For many in the audience, success is defined as the theatrical acceptance of stereoscopic 3D, that is to say, general availability of stereoscopic- enabled movie theaters showing full-length features produced by the big name studios. The current sales levels of stereoscopic products in the visualization, PC gaming, Internet or home video entertainment markets are either too small or not generally visible enough to be considered a success by many. A number of people expressed the opinion that a key factor that is limiting success is the low and inconsistent quality of 3D imagery. In particular, a need may exist for standards related to convergence and image separation leading to productions that can be comfortably viewed by audiences.

Daniel J. Sandin, University of Illinois/Chicago

Much of the content of the stereographics shown at the conference and much of the content of the discussion by the audience during the panel focused on theatrically released films in stereo, or more precisely, the lack of them. I suggest that computer-based interactive desktop stereo might provide a more fertile environment. Production in theatrically released films is a centralized hierarchical environment. Distribution for use on personal computers is an area where an individual or small group can make a difference. Computer games, for instance, represent an economic phenomenon that is as large as theatrically released films but is far less centralized. There are also a lot of application opportunities in the areas of scientific and industrial visualization in addition to home entertainment. Instead of thinking of Disney films, think small screen, think home computer, and think interactive.

Mark T. Bolas, Stanford University

It is instructive to go beyond the perceptual infatuation many have with stereoscopic imaging (including myself), and concentrate on how it can improve human interaction with computer-mediated environments. This requires the inclusion of our bodies when thinking about systems that use stereoscopic displays. If the scene says I am falling, but my body is staying still, it will be wrong. If I am focusing at 2 feet, but the imagery was shot at 20 feet, it will be wrong. If I turn my head, but the HMD imagery responds 120 milliseconds later, it will be wrong. The art is no longer about simply producing and displaying stereo images - we have succeeded at that. The art is now about how to use such images as part of a system perceptually rich in dimensions beyond sight, and with content that is cognitively meaningful.

Neil A. Dodgson, University of Cambridge

Lenny opened the panel with some comments about his own experience as an entrepreneur in the stereoscopic industry. He said that his customers often ask for stereoscopic displays that do not require glasses; there appears to be demand for autostereoscopic display systems that is not yet met by existing systems. There is also increasing interest in single-user stereo and in desktop stereo. Users want larger angles of view and interactive rather than passive content.

With regard to the conference itself, Lenny reminded us that he has been attending the conference since 1981 and said that, when he started attending, one would have thought that we would soon have a market for autostereoscopic displays based on vibrating mirrors. Technology moves on and the fact that some piece of technology works does not mean that it can be successfully exploited commercially. It remains to be seen whether the displays demonstrated this year enjoy commercial success. Lenny pointed out that color television took 20 years from its introduction to 50% market penetration, so it may take a considerable period of time for stereoscopic displays to be generally accepted in any given application area.

The panel session then moved on to comments from each panelist followed by a general discussion. The principal thought coming out of the discussion was that stereo is pointless if we have no content. Two companies represented at the session, nVidia and DDD, have developed software tools to address the lack of content.

nVidia developed a driver-level stereo interface for their graphics chips for shutter eyewear. All that it requires is a tiny bit of extra programming in the game software. There are now about 500 games that work in stereo (a couple of these were shown at the demonstration session). DDD has a system that can semi-automatically convert existing monoscopic content to multiview (this was also shown at the demonstration session).

As a counterbalance to these efforts, a member of the audience pointed out that the lack of compelling content is not the only problem: current autostereoscopic displays also lack resolution and cost a lot of money. One specific example of this was given by David Mark (of Mark Resources). He told us that he is inundated with requests to do stereo visualizations of scientific data. He said the interest and content is there, but they are waiting for the displays to become good enough for their visualization applications.

As a third factor, we must consider potential market size, in addition to technology and content. Visualization has a market of thousands of displays but entertainment has a potential market of hundreds of thousands of displays. We have sales and content for the former but not for the latter.

The bottom line seems to be that we can do stereo: now what can we do with it?