Issue: Volume: 24 Issue: 11 (November 2001)

immersed in art

By Barbara Robertson

To many people who create virtual reality applications, a "cave" is nothing more than a large, immersive display system. It gives automobile designers a way to display life-size virtual prototypes of new car designs. It lets architects tour factory floors that haven't been built and archeologists roam through ancient buildings that no longer exist. It offers scientists substantial views of such complex three-dimensional systems as molecular models, atmospheric systems, and galaxies.

Put in the hands of artists, however, a cave becomes an amorphous blank canvas that offers nothing less than the opportunity to develop an art form that is so young it has no design language or definition. Whereas architects, designers, and engineers look at digital versions of reality in caves, artists use caves to explore alternate views of reality that are otherwise impossible to see, and to create experiences otherwise impossible to imagine. In doing so, artists are beginning to define both a new art form and the ways in which people might communicate in the future.
Image courtesy Art Impact, Collective Retinal Memory by Maurice Benayoun.

"You can consider the art being created for caves to be speculative research in the human-computer interface," says Dan San din, co-director of the Electronic Visualization Laboratory (EVL) and professor in the School of Art and Design at the University of Illinois in Chicago (UIC). "The artists are creating extraordinary interactive experiences, which are completely unlike mouse and 'wimp' [windows, icon, mouse, pointing] interfaces. They don't have icons. They don't have mice. They don't necessarily have pointers." Instead, they have their bodies.

As some game players discovered in September at this year's Ars Electronica Festival in Linz, Austria, being immersed in life-size virtual reality while using their bodies as pointing devices can be a startling experience. During the festival, gamers could compete in a Quake "shoot out" using a version of the popular game ported to a 3- by 3- by 3-meter cave housed at the Ars Electronica Center, a five-story, interactive "museum of the future."
Each picture in the virtual Travel Tunnel has hidden parameters such as subject, colors, and era. As a viewer moves the images with a joystick, Z-A's Profiler software continually analyzes the action and seamlessly presents new images that reflect the

"There were two avatars, one representing each Quake player," says Horst Hortner, director of the center's R&D department, the FutureLab. "So instead of sitting in front of a monitor using a keyboard and mouse, the players had to jump around inside the cave. To move their avatars to the right, they had to move to the right." It sounds simple enough, but many Quake experts had trouble adapting. "It was very funny," Hortner says. "They'd say, 'Hey wait a minute, I don't have my keyboard, how do I do a crunch?' and I'd tell them, 'Just bend your knees.'"

The first cave was developed in 1991 and demonstrated in 1992 (at Siggraph) primarily by three people at EVL: Sandin, Carolina Cruz-Neira, now a professor at Iowa State University, and Tom deFanti, co-director of EVL and professor of Computer Information Science at UIC. It offers what Sandin calls the four pillars of virtual reality: a wide angle of view to immerse participants in images; stereo 3D, which puts a different image in each eye; interaction, to give participants the ability to move within the environment and to change it; and viewer-centered perspective, which is calculated in real time from a participant's point of view. "Viewer-centered perspective is the most interesting to me," says Sandin. "It's the first redefinition of perspective since the Renaissance, which gave us the camera's point of view."
In Margaret Dolinsky's "Strait Dope," participants discover their own path through a confrontational world of colors, characters, voices, and noises that operate under an alternate logic.
Image courtesy Margaret Dolinsky.

When people describe a cave, they do so in terms of its "walls," which are the display surfaces. Caves can have one wall, six walls, or even, potentially more. The classic cave is a so-called "four-wall" cave, a 10- by 10- by 9-foot "room," in which images are displayed on three walls and the floor; one wall is left open. Real-time, stereo 3D computer graphics images, usually generated on an SGI Onyx workstation that is often equipped with an Infinite Reality graphics engine, are typically beamed onto the four screens using Barco CRT projectors. Rear projectors put images on the walls so that groups of people walking around inside the cave don't cast shadows, and front projectors installed above the cave display images on the floor. In addition, many art projects incorporate sound.

Cave participants wear stereo 3D glasses, and their movements are tracked, usually via Ascension or Polhemus motion tracking devices. Typically, a small sensor attached to the glasses tracks head movement, creating the viewer-centered perspective, and often one or two trackers capture hand movement. In addition, a joystick or wand sometimes helps participants navigate through the virtual world.
The SAS cube, a 3- by 3-meter portable cave developed by Z-A, Barco, and ClartE, can function with two, four, or eight PCs running Windows 2000 or Linux, or with an SGI Onyx.
Image courtesy © SAS3

Unfortunately, the high cost of these Onyx-based systems has meant that most caves exist in research labs and in the engineering and science departments in universities, not in art museums or art departments. Thus, researchers have created viable alternatives such as ImmersaDesk, developed at EVL and licensed to FakeSpace, which also licensed EVL's original cave design. "The ImmersaDesk is one of the first devices we created as an attempt to get virtual reality art into galleries and museums," says Sandin. Although the device, which is in a drafting table format, has only one screen, Sandin notes that it's still immersive. "If you belly up to it, it fills your vision, and you don't see frame edges," he says. "But of course if you rotate your head, you're back in the real world."

More recently, Sandin and others at EVL have been looking at other one-wall alternatives that they hope will make it possible for museums to show cave art. And, during the Ars Electronica Festival, comparatively inexpensive PC-based caves shown by the FutureLab and by Z-A hold promise for more widely accessible caves in the future.

Today, though, only two publicly accessible museums in the world show cave art projects, NTT's Inter Communication Center (ICC) in Tokyo and the Ars Electronica Center. And while student cave art projects are developed at schools such as SUNY at Buffalo, the University of Indiana, the Fraunhofer Institute in Stuttgart, EVL, and others, the Ars Electronica Center is the only organization that commissions cave art projects.
Each head in Margaret Dolinsky's "Beat Box" represents an interval in a musical scale. Collectively, they form a virtual machine called a "Beat Box."
Image courtesy Margaret Dolinsky.

This year, the commissioned project was "Alive on the Grid," an experiment in "teleimmersion" among caves that was created by Sandin, his students, former students, and colleagues. Using software developed at EVL, six caves around the world were connected within one application that included eight different virtual worlds, each a unique interactive art project. "What ever happened in one copy of the world happened to the copy of that world in all places," Sandin says. "If I picked up an object in one place, the object moved in that world as seen in all the other places." In many of the worlds, cave participants were represented as avatars, and in addition to images, sounds were synchronized-the participants in different locations would hear the same sounds and could talk to each other.

In Sandin's "Looking for Water", for example, cave dwellers in Chicago showered avatars from Buda pest with digital water. With Margaret Dolinsky's "Beat Box," participants in Sweden collaborated with viewers in Indiana to create music by moving objects in a virtual world. In Josephine Anstey's "PAAPAB" world, a puppet manipulated in Linz could dance with a puppet being manipulated in Buffalo. Although people have long interacted with avatars in virtual worlds over the Internet, this was the first time that kind of interaction happened in fully immersive interactive environments where, for example, the avatars being controlled by people in one location could dance with any of 40 life-size avatars that were also being moved in real time by people in other countries.
The colored lines have the dynamic characteristics of water in this "Alive on the Grid" cave application by Dan Sandin, which was shown during the Ars Electronica Festival. People in six different locations in the world could shower each other's life-

Sandin and Hortner both noticed interesting social behaviors emerging among the participants. "It started during re hearsals and then be came pretty constant during the show," says Sandin. "People tended to hang out together. They would go to the place that linked to all the worlds, meet up with people there, and then they'd all jump off to some place together and play in that world."

Hortner also noticed that people would spontaneously help each other. "Some of the places had a one-wall cave, and some used a keyboard and mouse rather than a wand, so the people couldn't do everything that those in fully equipped caves could do," he says. "I saw people pick up objects and put them in the hands of those having trouble, without prior discussion."
"World Skin," an award-winning work of art by Maurice Benayoun and Jean-Baptiste Barrière, uses a cave to immerse people in images of war and offers them the option, as tourists on a photo-safari, to take pictures. The blank, white areas show where

In addition to the technical advances in VR projects, results such as these are what Hortner and others at the Ars Electronica Center look for in the projects they commission. "At the moment, virtual reality is at the same place in its development as film was when people were pointing one camera at a stage," says Hortner. He notes that when people simply put a theater experience onto film, the camera recorded a continuous unity of time and space. Film developed into a unique art form when film editing tools, for example, made it possible to cut from one view to another. When directors could change the unity of time and space, they created new paradigms for presenting visual information and telling stories-ones we are completely familiar with today.

"These paradigms have not yet been developed for virtual environments," Hortner says. "So what we look for in work we commission are projects that make intense use of the subjective space, that do something more than replicate reality, that drive VR a little bit further."
The Ars Electronic Center's FutureLab invited Viennese artist Peter Kogler to extend his ongoing experiments with architectural spaces into a cave to explore whether the visual tricks and illusions he uses in the real world would affect people in virt

Hortner points to the award-winning "World Skin," which was commissioned by Ars Electronica. "It is an art piece, not a cave application," Hortner says. "But it uses the cave and is only possible in the cave." In "World Skin"-created by Maurice Benayoun, founder of the production studio Z-A in Paris, and musician Jean-Baptiste Barriere-people wearing 3D glasses are in a cave filled with photographs and sounds of war. The photographs from WWII and Bosnia-women with crying children, soldiers standing and dying, tanks moving, and so forth-are texture-mapped onto planes, like billboards. One person acts as a driver through the battleground; everyone else is a tourist. Hanging from the ceiling are eight cameras. As the life-size images seem to move toward the participants in the cave, no matter which of the three walls they face, they begin taking snapshots, at first tentatively and then almost compulsively. Areas in a camera's view are highlighted in red. When a picture is snapped, they see a flash, the area is blanked out with white, and an intense sound erupts. The pictures are printed or posted to a Web site.

Writes Benayoun on his Web site, "With each click of the shutter, a part of the world is extinguished.... The rawest and most brutal realities are reduced to an emotional superficiality.... One 'takes' the picture, and the world 'proffers' itself as a theatrical event."
Participants inside "Uzume," a six-walled cave project created at the Fraunhofer Institute in Stuttgart, interact with swaying lines that seem alive. The lines are in a force field that is affected when the participant moves, and can be directly moved by

He says of the work, which was inspired in part by the realization that he took pictures to forget as well as remember, "I will never do something that in tense again. Emotionally speaking, it is not easy to reach such a level."

One of the first artists to work with virtual reality, Benayoun has now created several projects that consider how people interface with technology, virtual reality, and each other.

In the cave work "Crossing Talks" (presented at ICC in 1999), he demonstrates how telecommunication networks confuse our perception of proximity. On the cave walls are an infinite number of rooms with people talking in several languages. People in the cave can communicate to a few who are connected by the Internet, but only when the cave dwellers are in balance. If too many people move to one side of the cave, the VR world tilts like a raft, making communication impossible. Thus, the people in the cave must communicate in order to telecommunicate.

With "Art Impact," a project carried out over the Internet and in the Georges Pompidou Center in Paris last summer, Benayoun created what he calls a "collective retinal memory."

"I've been working a lot on leaving traces," he says. "We can easily imagine leaving footprints. But we don't leave cave prints." As people looked at ever-changing, 360-degree virtual reality images (similar to Quicktime VR) on the Internet and in VR binoculars at the Pompidou center, the images they were most interested in were projected onto a 4-by-11 meter screen at the Pompidou Center, creating a continuing blend of images from all the sources. Benayoun calls it painting with glances.
The woman's face inside this cave interacts with a visitor-the 3D head rotates to focus on the person walking about in the cave; when asked a question, she answers. Created by Catherine Ikam of Paris, "Face a Face," explores how people react to a hybr

In his first virtual reality piece, "Is God Flat?," people used a joystick to "dig" into a thick, three-dimensional brick wall on screen. As they dug, two-dimensional religious images would appear and float past them. Benayoun says the message of that work is not the illusory pictures of God, but the viewer's path, which built the architecture of the virtual world. "The VR world tells you that you exist, and because you exist it becomes different," he says.

Technically, he accomplished this work by creating the bricks in Softimage with layers of 3D brick textures, 128 pixels by 128 pixels by 128 layers. Viewers can dig in any direction through these cubes and see a brick effect. To describe how the path is created, he uses a sugar cube box analogy: "The box is empty when you start except for one cube," he explains. "Before somebody digging arrives at the end of the cube, another one comes just at the right time." By placing this cube where the digger would otherwise have popped out of the first cube, the digger can keep tunneling.

"I think what was wrong with interactive work in the early '90s was that everything worked with selection," Benayoun says. "You would have icons and lists of objects, and you could click on one of them. I didn't want people to choose. I wanted people to have an experience closer to the real world, where things are hidden and you have to do something to get them," he says.

He has extended the tunneling idea with projects that connected the Museum of Contemporary Art in Montreal with the Pompidou Center, and Paris with New Delhi. This time, Benayoun replaced the brick layers with pictures. To "dig," the people simply looked at an ever-changing collage of pictures onscreen, and while they did, software continually analyzed what they were looking at, the time they spent, and how quickly they moved the images and changed them accordingly. "The system analyzes what seems to seduce you and uses this information to select what it presents next," Benayoun says.

"Maurice [Benayoun] says virtual reality lets us explore things we can't explore in reality and research aspects of reality that are not accessible or possible to visualize," says Petra Gemeinboeck, an artist who has worked at Z-A and is now an MFA student at EVL.

While at the Fraunhofer Institute in Stuttgart, Gemeinboeck, an architect by training, collaborated with Roland Blach to create "Uzu me," a project for the Institute's six-walled cave, which has since also been shown in ICC's four-walled cave in Tokyo. In the six-walled cave (four walls, floor, and ceiling), white or softly colored lines move on a deep blue background. "Your presence influences a force field, creating waves," explains Gemeinboeck, "and with your arms you can affect the lines more directly." The lines exist in, and are moved by, the force fields and also by a user wearing two hand trackers and a head tracker. There is no other navigation in the virtual space, and the lines seem to have a mind of their own. "Some artists are interested in controlling what the user sees and does," Gemeinboeck says. "I'm interested in dynamic behaviors, the relationship between the person and the space, and the temporal quality of the space. I think that making things that look real reflects neither the potential nor the strength of this medium. I think the potential is to develop another reality."

Other artists explore the potential of VR in other ways. Al though Anstey, an EVL graduate who teaches at SUNY Buffalo, is interested in continuing the "PAAPAB" world created for "Alive on the Grid" by adding more puppets and flocking behaviors, her real interest is in narrative-based projects such as "The Thing Growing" (See "Interactive Fiction, pg.34, April 2001) and in the emotional and dramatic possibilities of the medium. "I want to push the intelligence and the reaction of the system to users, to store information about what users do and choose," she says.
An interior stream of consciousness formed the painterly cave environment in Margaret Dolinsky's "Blue Window Pane," shown at the Ars Electronica Center and at ICC. Participants explore a mentally confrontational world by moving to the next image that

Dolinsky, also an EVL graduate, who teaches fine art at the University of Indiana, likes to create subversive confrontations and experiential metaphors with in virtual environments such as the cave and so-called Immersa Grams. (Immersa Grams, snap shots of virtual reality taken inside a cave that are viewed in light boxes, are created using a process developed at EVL and at (art)n in Chicago.)

In a cave work called "Strait Dope," for example, Dolinsky sends things flying through the air at people. "I was experimenting in stream of consciousness movement," she says. "Some people thought it was frightening, and others enjoyed it." In "Blue Window Pane," which has been shown at ICC and is installed at the Ars Electronica Center, a viewer is confronted by a woman's face chanting, by faces staring as the viewer climbs stairs, and by sounds of people whispering and rambling. "This field is so young that people are exploring it in different ways," she says. "There is no concrete language."

For his part, Sandin explores dynamic movement using physically based simulations of natural elements such as water and air currents in caves.

"Almost every application gives us a little letter in a design alphabet for this new venue," says Hortner. One of the keys to creating that alphabet, however, will be in getting the tools into the hands of more artists, and the artists' work into more publicly accessible venues.

To that end, in September the FutureLab introduced the Ars Box, a PC-based cave equipped with Nvidia's GeForce graphics cards, which can produce stereo 3D images and can run most cave applications. It supports CAVElib, which controls the devices in a cave, OpenGL, and Silicon Graphics' Performer, all running on Linux. With the Ars Box, museums could install a cave for a fraction of the price of an Onyx-based system yet still run most of the content already created for caves.

Similarly, Z-A introduced its SAS cube, a portable, PC-based cave developed by David Nahon, which is OpenGL-based and runs on both Windows 2000 and Linux. Benayoun explains that in French, a "sas," is the room that leads from one place to another. "They've solved the problem of programming PCs in a clustered ap proach," says Pape of the SAS cube. "It's good to see."

And Sandin has been working with Gregory Dawe at EVL on something they call a Try Desk, three screens arranged in a pod. It would have the cost and installation advantages of one-wall caves yet provide some of the social interaction common to four-wall caves.

Meanwhile, Dave Pape, who inherited work on the CAVElib from Cruz-Neira, has been working on a scripting language for caves. Named Ygdrasil (YG) after a Norse god, it gives artists a way to create three-dimensional worlds for the cave and specify behaviors and interactions in those worlds, without having to learn C++ or depend so heavily on programmers.

With the cost of devices dropping, cave art may finally be on the verge of becoming more accessible to artists and more visible in schools, galleries, and museums. "If so, then we will finally begin seeing more applications," says Pape, who agrees with Sandin that this is important. "Artists, filmmakers, and other people need to learn what's possible."

Barbara Robertson is Senior Editor, West Coast, for Computer Graphics World.