(A somewhat longer version is available at http://www.gsd.harvard.edu/~servin/vrlandscape/long-version.html ).
Imagine a meeting between a landscape architect and some clients or colleagues, in which a proposed new landscape -- designed using all state of the art CAD and GIS software -- was presented in such a way that any of the participants could grab the mouse, or joystick, and take the group to a particular viewpoint, or along a particular path; could click on a proposed planting or structure and get an instant listing of details, cost, and maintenance implications; or look around in a 360-degree panorama from anywhere. Imagine, even, that all this was being presented in life-size, sense-surrounding projection, and not just on a flat computer screen. This is the promise of 'Virtual Landscapes' -- or Virtual Reality (VR) applied to landscape design. And even though some of this still seems to be the stuff of science fiction, all of the technologies necessary to realize the above scenarios already exist -- and are being used.
The common meaning of 'virtual' in computer contexts is just 'simulated' : 'virtual memory', for example, is computer memory which acts like RAM, but is not really provided by RAM, rather is simulated by using hard disk storage. And so 'Virtual Reality' is just simulated reality, and a 'virtual garden' is nothing more than a simulated garden; simulated in a computer, using CAD, GIS, image processing or animation, for example.
If 'simulated' is all it takes to be 'virtual', then any computer representation -- CAD drawing, GIS map, digital image-processed photograph -- qualifies. Indeed, you can find a CD-ROM promising 'Virtual Landscapes' which contains striking images of landscapes ('Rocky Mountain Peeks', see http://www.csn.net/malls/rmdp/vldesc.html ); you can find a web site called the 'Virtual Landscape and Garden Center', an on-line shopping and information center for gardeners of the real kind (at http://www.farwestgardens.com/vgarden.htm ); you can find some 'virtual garden tours', for example on the web at the "Grow Zone", (www.plants.org)( http://www.plants.org/cultinterests/tours/index.html ), where you can virtually visit a variety of gardens around thew world, including the Japanese Garden at the Missouri Botanical Garden (see Figure 1) at http://www.mobot.org/mobot/gardens/japanese/ ).
Figure 1 - the Japanese Garden at the Missouri Botanical Garden
Serious proponents of VR, however, offer three criteria for differentiating the 'truly virtual' from the simply simulated:
A recent project was mounted at the Museum of Art & Architecture in Frankfurt , Germany, on the occasion of a 200 year centennial exhibit of the landscape of the 'GartenReich', in the province of Sachsen-Anhalt (in what was Eastern Germany). For the occasion of the exhibit, multimedia experts at the Institute for New Media in Frankfurt created a 'virtual model' of the garden, using a Silicon Graphics Inc. (SGI) computer to combine a three-dimensional model and detailed photographic images, which were used as 'texture maps' on the 3d model. This model was accompanied by textual descriptions of the garden elements, such as materials, date of construction, and historic commentary.
For the exhibit, a large auditorium in the museum was equipped with a rear-screen projector driven by the SGI computer, filling one wall of the room (a display approximately three meters by four meters). Mounted in the front center of the room was a wooden walking stick, mounted on a mechanical 'joy-stick' or three-dimensional mouse. Viewers at the exhibit could walk to the front, grasp the walking stick and use simple hand motions to 'walk' along the path, pausing whenever they desired, rotating their direction of view around 360 degrees, and by clicking on a built-in button in the walking stick, get the detailed textual information in a 'pop-up' window whenever one of the built elements was highlighted on the screen. In this way, museum visitors who could not get to the garden itself, could have the 'virtual experience' of it. Figure 2 shows images from the 'Virtual Woerlitz' simulation. More images, information and a down-loadable movie can all be found on the Web at http://www.inm.de/projects/woerlitz.
This project met all the criteria for a virtual landscape: 'immersive', because of the large screen projection, in which the landscape is in scale with the human body (that is, larger, rather than represented on a small screen); 'navigable', because of the ability of the viewer to control the motion, by the 'walking stick', and by turining the view around 360-degrees; and 'interactive' because of the built-in 'hot buttons' in the landscape, which revealed more information when clicked upon. At the same time, it was still a primitive example: the immersion was not complete (no more than an ordinary movie theater), the navigability was limited to a single pre-determined path; and the interactivity was limited to a few buttons on a few items in the landscape. Many video-game arcades now offer more of all of these aspects of VR. Nonetheless, 'Virtual Woerlitz' was a novel and rather exhilirating experience.
While the Woerlitz projected used a high-powered SGI computer, similar effects are now readily achieved using ordinary desktop computers. A recent project sponsored by the English Heritage society, the computer company Intel, and the company VR Solutions Limited, has made available a 'virtual Stonehenge' (See figure 3). The model, and the software required to view it (available only for Windows 95 and Windows NT), are available on the web at http://www.superscape.com/intel/shenge.htm . With this model you can stand beneath the giant stones, whose texture and markings have been accurately and realistically reproduced, and watch the the sun rise, or 'walk' around the circle while reading about the history and lore of the paleolith. Since the public is no longer permitted to walk among the stones, this virtual model is now the only way to experience the space created by these giant stones, or their textures, or the experience of a sunrise or sunset thorugh the astronomically aligned arches.
(See also http://www2.ucsc.edu/people/trillian/stonehenge/start.html for more information about Virtual Stonehenge on the web.)
Figure 3 - Virtual Sunset at Stonehenge.
An important piece of both these examples is the use
of photographs to give 'realism' to the virtual scene. The technology
called 'QuickTime' (developed by Apple computer for animations and
multi-media) has been used to develop an extension called 'QTVR'
(QuickTime Virtual Reality), in which two-dimensional digital
photographs are 'wrapped' on three-dimnsional surfaces, such as the
walls of a hall, or room, or on a cylindrical background. Special
software (a QTVR Browser) is used to navigate through the virtual
space, so that the viewer's onscreen image changes under
mouse-control. The technology required to seamlessly merge panoramic
photographs into a three-dimensional simulation involves a special
camera/tripod set-up and some special software, but has been used for
a variety of practical purposes by landscape architects and
planners.
In St. Paul, Minnesota, the department of Parks and Recreation used QTVR technology to help illustrate concepts for a master plan for the historic Como Park. Planners and designers, as well as the public were able to 'walk' the park, in both existing and proposed conditions, from a CD-ROM produced by a local media production company.
See http://www.stpaul.gov/welcome/ for a sampling of other QTVR views of the city.
In another approach to VR, the environmental group No Ka Oi Foundation, for their "S.A.V.E. Maui" project, offers a virtual tour of the Hawaiian Island of Maui, at http://nko.org using the VRML software called 'Cosmo Player', from Silicon Graphics Inc (for more information, including a download copy of the softwarer and a vast library of VRML information see http://vrml.sgi.com ). You can 'fly' over and around a three-dimensional model of the island, dotted with three-dimensional 'icons' (colored spheres, pyramids, cubes), each of which leads to the WWW website of some institution or organization -- schools, museums, the astronomical observatory, and others. In this way, the VR landscape model serves as an interface to a large space of related websites, and combines visual interface with real geography in an intriguing way.
Figure 7 - Virtual Maui
Of all the three characteristics of real 'VR' cited at the beginning, 'immersion' remains still the most elusive because of technical limitations and expense. The 'head-mounted displays (HDM's) of science-fiction are not yet well developed enough for routine use outside of video parlors; instead, several approaches to more full-scale immersion are being tried. Rodney Hoinkes, for example, known to many in the landscape/computing circles for his contributions to the Center for Landscape Research at the University of Toronto, including the Polytrim software and the WWW Virtual Library of Landscape Architecture website, has begun Immersion Studios, a new commercial enterprise focusing in this area, on the web at http://www.imm-studios.com . Figure 9. shows a schematic of a 360-degree 'immersive theater', in which a virtual landscape can be presented, by six (or in some versions, twelve) simultaneous projectors, each run by a powerful computer and linked toigether to project an immersive 360-degree virtual landscape.
Figure 9 - a 360-degree 'immersive theater', from Immersion Studios
Rodney Hoinkes says of these technologies:
"We see tremendous commercial applications of the techniques and
technologies we have developed, such as those used in the 'Crossings'
project, for immersive and interactive environments on the large scale.
We are creating real spaces in which social interaction leads and follows
from virtual interaction and virtual spaces. How these developments will
relate to the practice of landscape architecture, or education, we have
yet to see. But clearly the potential is vast."
Another approach is taken by the the Finnish company 'Plustech Oy' who have developed some virtual technology to be used in conjunction with forestry equipment. Their 'walking machine' is an advanced forest operation machine that uses articulated legs -- like a giant insect -- instead of wheels or tractor-treads, and so (the company claims) can cause far less environmental degradation in forestry operations such as timber extraction. The company has developed a hardware/software combination they call the 'landscape simulator', which combines two different pieces of software and some special-purpose hardware to train foresters in the use of their special equipment. One part of the simulator uses GIS and CAD data to generate three-dimensional perspective views of forests and cutting operations, enabling the user to preview the visual impact of different cutting schemes. The other part uses a set of real controls from the equipment, mounted to a special chair and connected to a computer, to train operators of the 'walking machine'. Figure 8 shows an image of the landscape simulator, with an operator seated in front of a simulated forest. (More information can be found at http://www.plustech.fi/simut.htm)
Figure 8 - Virtual Forest and Forestry Machine from Plustech Oy
Some of the above projects, like the last two, are rather far from the routine domain of the profession of landscape architecture. Some of the other cited ones were completely embedded in the demands of practice. As with all other developments in computer technologies, a wide range of kinds and levels of experiments are necessary to flush out the promise and the real applications of VR. Some landscape architects may be naturally wary of VR -- after all, a virtual garden will never be a real one. And the oft-cited concern of 'basking in the glow of the computer screen', rather than on a real beach with real sunshine, is worrisome indeed, if a generation of humans come to know gardens, or the landscape, only through these virtual representations. But this genie is already well out of the bottle, and will not be stuffed back in. A whole new generation of landscape architects is being trained with the expectation of VR in their grasp in the next decade.
It's unclear exactly all the ways VR will find its way into landscape
architectural design, but the opportunities are many. The interactive walk-through scenario presented at the beginning, in many variations, is already a real possibility. The possibility of performing research in 'virtual', rather than 'real', environments is also intriguing. And the
promise of multi-user, multi-participant virtual environments opens up new
possibilities which will be quite un-like traditional conferences or community meetings.
The proliferation of examples already
being tried, and the increasingly affordable availability of high
powered computers and software, suggests that before the year 2000 we
will have more than a few legitimate examples of 'virtual
landscapes'. See you there!
Stephen M Ervin
Lecturer in Landscape Architecture
Director of Computer Resources
Graduate School of Design
Harvard University
email: servin@gsd.harvard.edu