The difference between virtual reality and the real world is that in the real world anything can happen, but usually doesn't, and in the virtual world there is only the illusion that anything can happen. So, in a virtual world, especially one that we've paid for, we want some action a sassy bartender, a mysterious dame. And the problem with virtual reality is formulating the action, without turning the experience into a one-track flight path that negates all the interactivity that virtual technology seems to offer.
Currently, much of the research effort in VR is targeted at getting high quality sensory substitution to eyes and ears and skin in real time, and into good real time control for users in 3D environments. These are hardware-software problems which are slowly but surely resolving themselves. Silicon Graphics have a brand new sales and development site in Reading, housing the Reality Centre; a small theatre capable of playing some of the most powerful real time simulations on the planet. To define the current state-of-the-art, here are two examples of their recent work:A virtual model of a real oil rig, built from the original working computer model drawings. The detail is so fine you can see the flats on individual nuts and bolts and the knurling on the pipe control wheels and the display rate drops to a jerky 3-5 frames per second, because of the processing overhead. At such a level of detail and so many objects, even the Onyx Reality Engine2 is struggling.
An underwater simulation, where the graphical fishy inhabitants have limited embedded intelligence; they can shoal, detect threats, and avoid obstacles. They even have predators, virtual sharks that swim in a random pattern, morphing tail fin movements instead of animating pre-set sequences.
How long before developments in parallel processing and 3D graphics hardware put virtual sharks on our desks or in them? Assuming that steady improvement in computing power continues, if VR is going to fulfil its promise, where is it going to go, and what pointers do we currently have to get there?
The latest batch of 3D computer games making their way into the marketplace for Xmas 1994 are a foretaste of things to come. The very real limitations that 3D places on current desktop hardware is forcing designers into shortcuts and workarounds that spin off strange new variants and themes. For example, the Psygnosis game Ecstatica is a 3D horror adventure in a strange cartoon style that quite belies the demons and the gore; a little like a cross between Hammer Horror and The Magic Roundabout. It's a good game, but it clearly reveals the limitation of animating 3D images in real time on affordable current equipment. Another workaround is the much-hyped "interactive movie" on CD-ROM, where the viewer can follow through a limited number of scenarios in very high quality environments, mixed from pre-rendered 3D computer graphics and pre-recorded digital video and sound. Sierra's forthcoming game Phantasmagoria is also a horror fantasy, but in pre-recorded, almost photo-realistic 3D detail. US Gold market Under a Killing Moon, a detective fiction game which comes on 3 CD-ROMS nearly two gigabytes of memory. These games will look and sound beautiful, but only because you see and hear what has been pre-recorded. The "interactive" part of interactive movies is limited to the director's choices the player just chooses the sequences of recordings to view. This is not the same kind of interactivity that hallmarks virtual reality, where the players can, in theory, wander at will, and compete and collaborate with each other, as well as the program. The choices that can be imposed on players in virtual worlds must be altogether more subtle than the forcing of paths through a pre-scripted story. But there are other games around that offer more of a window on virtual worlds.
All of a sudden, the fashion in computer games is multi-player. Flight simulators like Air Warrior can be played networked by modem and suitable bulletin board software: other games offer both modem and cable network support. Bullfrog's Magic Carpet gives you two player interaction. Doom from id Software gives you up to four player collaborative play or bloody DeathMatch, and their much-awaited forthcoming game Quake will offer up to ten players the chance to knock seven kinds of mayhem out of each other. Stuff one of the new PC-compatible virtual headsets from VictorMaxx into the graphics connector of your Intel Pentium-powered machine, and you could be virtually interacting with the best, indeed, short of spending twenty times as much money, this is the nearest most of us will get to virtual reality in 1995. But how much interaction are you actually getting in these games? Stimulating they may be, but all that these games are simulating are variations on MOVE, SHOOT, OPEN, PICK-UP and DIE and, in fact, DIE is a function of someone (or something) else's attack on you. True, there may be a few puzzles to solve along the way, but thematically, these games are not a great deal different from text-based adventure games from ten years ago. So is this really all there is to VR entertainment? Not entirely.
Brenda Laurel and Rachel Strickland have built Placeholder, a virtual world design research project for Interval Research, USA. Placeholder is very different from the current generation of simulated combat. Set in a tranquil river valley, a waterfall and a cave based on real places in the Canadian Banff National Park, the player can take the part of Crow, Snake, Spider or Fish and in doing so take on the viewpoint, the movement and the voice of their chosen creature. The critters become a kind of "smart costume", altering both the appearance and the behaviour of the wearer. Just as real travellers can leave physical evidence of their passing, Placeholder players can leave "Voicemark" sound bites for others to find, and search their way through the trail laid by previous players, guided by pre-recorded Voices that seemed to emanate from inside their heads. The virtual landscape accumulates definition as time passes. It is a quiet, strange experience, devised and designed as much from theatre as from high technology and hopefully a pointer to VR designers for the future.
Meanwhile, other media and leisure industries have been quick to cash in on VR. Television has been capitalising on youthful interest in new technology with a variety of programmes and game shows purporting to be virtual in some way. Like many computer games, the claim to "virtual reality" usually interprets as "there's some real time 3D graphics involved, honest..." But watching game participants competing in Timebusters (BBC2) or Virtually Impossible (ITV) is only interactive for the players, not the audience; a long way from real VR. Still, Saturday mornings do have Ratz the Cat and RatRun, remarkable for the real-time puppeteering of a 3D animated cartoon head by a real actor, using a facial waldo to control head, mouth and eye movements. Expect to see such technology extended to allow actors to become virtual puppeteers for whole body virtual puppets. Though it may seem cheaper to try and make the creatures in virtual worlds entirely software driven, it is not a trivial task to program the artificial intelligence to convince you they are human. Sharks, perhaps, people.... Not just yet.
Leisure parks and arcades see VR as a cost-saving and safety-effective technology, an alternative to roller-coasters and a supplier of endless new ride environments for the cost of the software design. But the same issues of cost and system limitations that affect computer games affect the leisure industry. As public perception develops and demand grows, the simple flight simulators and boxing rings lose their appeal. The next ten years will see greater stabilisation of VR technology and greater user demand for realism. For companies like Virtuality and Battletech, who have pioneered the VR arcade game, the holiday is over. Last year, when Virtuality were W Industries, they had already recognised the problems inherent in providing an interface to virtual reality arcade interaction that was intuitive enough for players to learn quickly. If a player feels they are wasting many pounds or dollars just to learn how to use the machine, they are unlikely to persevere if they can't make some headway fast. Yet very few of us have any experience of interacting in 3D with the limitations imposed by the first flush of arcade VR - heavy headsets, clumsy, slow-moving graphics, cables hanging down our backs. The second generation of hardware is on the way, for sure. What remains to be seen is if the compromises and workarounds needed now and in the next few years to compensate for limitations in hardware spill over into the systems of the future - just as the limitations of text-based and 2D "windows"-style interaction is beginning to hold us back from fully 3D interface design.
Few parents are going to be happy to let young children experience immersive virtual reality. Indeed, when you are four or five years old, you're only just getting to grips with the real world, let alone a multiplicity of virtual ones. Instead, Veronica Pantilides, co-director of the VR and Education Laboratory, East Carolina University, USA, thinks that when schools can finally afford low-cost VR systems, they will be desktop VR or CAB systems. CAB VR uses screens and mirrors to place an apparently 3D image "inside" a box, within which the player can interact. The box can be sited on a desk, like Myron Kreuger's VideoDesk, or the players can be interacting with a giant screen in a room. Further in the future, she believes, we will see virtual environments for education merging with entertainment and mirroring the "Holodeck" of Star Trek The Next Generation. Both technologies offer school children new ways of learning.
Two broad areas of use in education look like converging: the use of the Internet to trawl vast amounts of data at sites world-wide, sharing chat and information with virtual classmates, and the use of stand-alone educational software or multimedia in the classroom. Internet use in the schoolroom is picking up pace in the USA and Canada, where the wealthier and more switched-on schools have climbed aboard Clinton and Gore's Information SuperHighway. As protocols fall into place for networked 3D environments, shared virtual spaces will evolve for use in education: Schoolchildren could take part in a virtual space launch at NASA, or visit a Jurassic Forest. Or imagine a virtual version of Shakespeare's London, peopled by a team of professional actors puppeteering the virtual inhabitants. Visit Shakespeare's Globe Theatre and see a performance of Hamlet as it might have been, fist fights in the audience and all, rather than the sanitised versions they might now get taken to at a local rep. Best of all, they could share their experiences with other children from all over the planet. But it will be a while before even limited access to global information resources falls into place in the UK. Here, things couldn't be worse. Most primary schools are badly under-resourced: many still have only IBM-style RM Nimbus PCs which only run RM software, typically five to ten years out of date. If they are lucky, they may have an ageing Apple Mac or two. Secondary schools might be lucky enough to have some faster PCs, and some CD-ROM drives, but without sound investment in IT across the board, the use of computers in British primary and secondary education lags far behind many their European and American counterparts.
Meanwhile, if resources are going to be shared in virtual spaces, some legal consideration must be given to issues like copyright, access rights, and payments for use, perhaps through some simple mechanism of automatic credit and debit for organisations that make information available for use, and who in turn use other resources. Furthermore, the co-operative environments envisaged by Veronica Pantilides create a less competitive and test-oriented way of working for children, perhaps more suited to the modern world, but out of keeping with "no-cheating-get-on-with-your-own-work" traditionalists. Nolan Bushnell, one of the founders of Atari, once described video games as a training ground for virtual reality. Research into the effects of Nintendo on the youth of America showed that along with side effects like repetitive strain injury went the development of problem-solving, pattern recognition, resource management, logistics, mapping and reasoning skills, memory testing and collaborative abilities. Not bad for a video game.
No peek into the future of virtual worlds can ignore sex. Every new media attracts its pornographers. Gutenberg's printing press wasn't just used for churning out bibles, and the flood of naked ladies that engulfed Europe and America in the wake of cheap mass photography was just a prelude to further reproductive technologies: home movies, the video camera, computer graphics and animation. Software author Mike Saenz (of MacPlayMate, Virtual Valerie and DonnaMatrix fame) has always claimed that easy distribution of erotica spurred the uptake of video in the seventies and eighties, an argument that is often echoed but seldom verified. Many are now making the same claim about the uptake of CD-ROM in the 1990s and if the mushrooming CD porn booths at big computer trade fairs like Comdex is any indication of a trend, then the pundits are right. And virtual reality porn? Well, the truth of it is, apart from a few star turns, it has yet to come into being. ImageX's ZyberFantasy in New York City is little more than a virtual video booth with a head mount display, and a recent experimental system for ISDN telephone sado-masochism in Cologne is reputedly under investigation by the German police. One of the few UK-based experts on virtual reality for sex is Trudy Barber, who says: "Serious VR sex is sadly lacking in money or research. It's full of seedy back streets and lack of interest. But it is going to be a vital part of VR. If VR allows people to take control of their fantasies, then it seems absurd to deny them sexual fantasy. In VR you can be multi-gendered ambisextrous. If you can penetrate yourself, why bother with anyone else? But VR won't just be masturbation - if you are working away from home, you can still keep a relationship alive, virtually. It changes the way people think about themselves."
Each day brings new solutions to bear on the problems of real VR, and new problems emerge for its actual use. Already there are plans for a 3D mark-up language for the Internet, a kind of virtual version of HTML (Hyper Text Mark-up Language) used to create World Wide Web Pages in browsers like Netscape and Mosaic. There were two NewsNet announcements in the first week of December of freely distributable 3D world viewers, Vignette from Cybertech, USA, and a 3D Viewer from True-D Software, Oxford, UK. All this bring distributed virtual environments on the Internet a little closer, and forces the hands of those companies like Renderware or RenderMorphics, who are struggling for control over general standards for 3D systems. But games like Doom, also given away free and with an estimated million plus players world-wide, blow the gaff on issues of violence and parental control - how realistic does murder (even of mutants) have to be before it is beyond the pale? And why, for that matter, should blowing up aeroplanes or spaceships be somehow easier to sanction than gory death? And there are physiological issues to consider, too. Just what is the effect of staring at a screen two inches in front of your eyeballs for hours on end? How long does it take to recover from the disorientation caused by the illusion of weightless flight, if you are driving home?
But just as there are threats, there are the promises of benefits as well. Virtual entertainment and education offers a future hope to the disabled in the form of truly level playing fields. It offers the reconstruction of lost buildings and lost times, of the safe training for dangerous tasks and sports. Sadly, most of this is still a long way in the future for all but the richest of us. If you can afford £5000 a year to upgrade your computer hardware and keep up with the technology, you might stand a chance of playing out the latest in home VR as it reaches the shops. If you can afford £1600 plus a year for your ISDN line, you may have a chance of accessing networked VR when it reaches the Internet. If you are a student with cheap access to computing resources, then you are very, very lucky. But if virtual reality is to make any impact, it must quickly go beyond the mute examination of wonderful texture mapped graphics and the zapping of hordes of quasi-intelligent but homicidal aliens. We have to have systems that support speech, that are easy to navigate, and where intelligence is not just about homing missiles. We have to have systems that will cheaply and easily support more than one person at a time, and we have to have access to a general communications infrastructure to enable that, not a hotchpotch of satellite, cable TV and ageing twisted copper telephone wires. The introduction of virtual reality is going to require as much political willpower and artistic input as technological leverage to make it happen. As the ancient Chinese curse goes, "may you live in interesting times".
Ivor Benjamin is a systems analyst and a theatre director. He is conducting research into interactive virtual reality systems at the Centre for Human-Computer Interaction, City University, London. You can contact him by e-mail at ivorb@cix.compulink.co.uk ivorb@cix.co.uk or i.d.benjamin@city.ac.uk
Back to PUBLICATIONS
HOMEPAGE