Digital visualizations of architecture are becoming more and more refined, and stereoscopic vision has been in use for many years in computer games, research, medicine and technology. “Virtual reality” (VR), however, implies something more: that the observer can also exert an influence on the images depicted. Is this technology applicable in architecture?

Since the main strength of VR lies in rendering sequences of movement and not in visual quality, the graphic depiction is often quite coarse. The technology plays an important role in the aircraft and automobile industries, however. Harald Hechler in the BMW research and information centre in Munich says: “With our high-end visualization equipment, decision-makers can select their own position in a predefined realm. The operation is extremely simple. The user can concentrate entirely on the scene before him and is not confronted with the complex technology behind it all.” By means of infrared technology, the virtual vehicle can be turned and tipped with a single navigational movement (ills. 1, 5), and by pulling the 3D model towards you, you can put yourself in the passenger cell again, navigate your way past the headrests to the driver’s seat and look around the car at will with the same angle of view as in reality (ill. 4). Jens Neumann of the Fraunhofer Institute for Production Plant and Construction Technology in Berlin defines virtual reality as a combination of three factors: immersion, interaction and imagination. Immersion means the entry of the observer into a scene as completely as possible.

Large planar projection surfaces, socalled “Powerwalls”, offer a broad close-up field of vision. “Caves” are a further means of obtaining a high degree of immersion. These consist of 3 ˜ 3 ˜ 3 m cabins in which three walls, the floor and the ceiling are rear-lighted. Caves are used mainly for simulations of seated activities. Although “workbenches” are less suited to a process of immersion in virtual space, they can be used for viewing objects like virtual architectural models. Workbenches consist of rear-lighted tables, either in the form of a purely horizontal surface or with an additional vertical wall at the back. Interaction implies scope for influencing the course of events in a particular scene and spontaneously intervening in real time. The space mouse, data glove, data finger and 3D joystick are various means of personal input.

In order to react to the input of a freely moving observer in a real-life manner, the computer must be able to determine the person’s precise position in real time. Finally, imagination is necessary to interpret the individual stimuli in their entirety as “reality”. If the technology also addresses the movements and sensations of the observer, the boundary between virtual and true reality becomes quite tenuous. Three-dimensional stereo visualizations are already used in architecture. At the Institute for CAAD at the University of Technology in Munich, an auditorium equipped with polarization-filter plant has been taken into operation to present rendered animations to students in cinema-screen format. Stereoscopic presentations are also used to great public effect in exhibitions and trade fairs, usually in the form of digital reconstructions of historical buildings and cities. Initial applications of virtual reality have already taken place in architecture.

In the summer term 2006 at the Leibniz Computer Centre in Garching near Munich, art students tried out the 3D pavilions they had designed on the internal VR installation. They were able to turn, tilt and even draw the rendered data model overhead like a physical object on the holobench so as to judge the quality of the internal space. Common applications of VR in architecture lie mainly in areas like the construction of factories, where one is concerned whether robot arms are likely to collide or whether bulky machinery can be brought into a building through a particular opening. In the future, the use of virtual reality can be extended as a digital aid in architecture. A building is spatially far more complex than a car, of course, so that one will rarely simulate the entire structure. The application will more likely be confined to a specific realm; but in the case of large projects, if the visualization helps to convince a client or the public of the merits of the design or to allay the fears of local residents, the outlay could be worthwhile.