ARCH'IT interview / Questioning the architecture of interconnected virtual environments

	CITYCLUSTER "From the Renaissance to the Megabyte Networking Age" is an installation designed by Franz Fischnaller and realized in February with a real time connection between Florence and Chicago. The project consists of the networking of two access points -the first being the CAVE at the Electronic Visualization Lab in Chicago, the other in Florence through the AGAVE visualization system- to the virtual shared terrain –made up of two VR environments: Florence, the "Renaissance Age" and Chicago, representing the "Gigabits Age". Visitors in tele-immersion can surf around and meet each-other, or exchange buildings and objects and ideas, creating their own ideal city. Franz Fischnaller is the founder of F.A.B.R.I.CATORS http://www.fabricat.com, an interdisciplinary group aiming to combine in an effective manner technology, communication, architecture, art and design. At the present time, he teaches at the Electronic Visualization Lab of the School of Art and Design at The University of Illinois at Chicago. [MLP]
[in italiano]	MARIALUISA PALUMBO: If you agree, I'd like to focus this conversation on the more technical aspects of the installation. In fact, going beyond the aesthetic and linguistic choices, I believe that trying to understand the basic material and non-material elements of this virtual architecture could be of great interests to architects. Regarding the general concept about which you wrote: "The system has been designed to produce an integrated computing facility and to implement a creative high-tech container in which multiple environments may coexist and be interconnected within a common, virtual territory". Could you please explain to me your definition of "integrated computing facility"?	[13may2003]
	FRANZ FISCHNALLER: Clusters, based on basic commodity components: hardware, software tools and applications that through their synergy can become a dominant platform with a multitasking operating system where multiple programs (applications) can be running at the same time. MARIALUISA PALUMBO: What does it mean that multiple environments may coexist and be interconnected in one virtual territory? FRANZ FISCHNALLER: I have a virtual-reality networking matrix were I host a collection of virtual environments: virtual ambiences, habitats, cities both real and imagined which can be interconnected between themselves. MARIALUISA PALUMBO: And what about if a new node hosting his own virtual environment, coded in the same or in a compatible language, would be added to the net? Would it be possible to the other users 'to access' this new virtual environment? FRANZ FISCHNALLER: If he has our software, the suitable hardware and tool interfaces, the permission to access to the virtual environment and the right networked speed connection, yes. MARIALUISA PALUMBO: About the software, what are the specific issues of a multi-user interactive virtual environment?
	FRANZ FISCHNALLER: "Ygdrasil", created by Dave Pape, is the software utilized for developing the project CITYCLUSTER. Ygdrasil is a framework developed as a tool for creating networked virtual environments. It is focused on building the behaviours of virtual objects from re-usable components, and on sharing the state of an environment through a distributed scene graph mechanism. Ygdrasil is built in C++, around SGI's IRIS Performer visual simulation toolkit and the CAVERNsoft G2 library. Performer provides a hierarchical representation of the virtual world database, called a scene graph. The scene graph is a tree that encodes the grouping of objects and nesting of 3D transformations, and provides tools for operations such as switching elements on or off. The project calls for the ability to "capture" elements from one city and add them to the other city. This poses a challenge because what constitutes an "element" may not be easily added to another geographic location elegantly. The elements envisioned range from entire buildings to simple architectural components such as windows and door treatments. Furthermore, the possibility of both permanent and temporary modifications to building textures was envisioned for the project. A collection of components was added to the system that allows the real-time modification of textures associated with models. These modules allow users to modify a texture at any specific location. This technique might be used to make textures transparent, thereby allowing the user to see through a given structure. This process can be expensive with regards to system resources and can give unpredictable results when used indiscriminately. Consequently, locations are limited in order to maintain close control on the nature of the models affected by this technique. Any attempt to immerse the user in a large complicated environment such as a city requires a highly optimized system. A strategy of replacing models having a high number of polygons with lower resolution models ("level-of-detail") will be used to reduce the load on the graphic hardware when rendering buildings and other objects that are not close to the user or cannot be seen from within the current context. In a multi-user environment, all potentially rendered models are loaded into the memory and excluded from the rendering process until they are required. Most geometry in the project will be static geometry that appears in the same manner for each user. Any dynamic changes to these models will use a set of tools that the lead technical advisor developed last fall. It allows for the material properties, textures, and vertex coordinates of models to be changed in real-time. Although these changes must be propagated to each client site, they are performed through simple parameter changes such as material color, texture position, and geometric interpolation position. MARIALUISA PALUMBO: A specific tool of the interface modality you designed is the "Meta-Net Page", a kind of window or multifunctional view panel inside the main scene. The system is extremely interesting (even if I found it a bit hard to use!); can you tell me how it works technically? How much importance did you give it with respect to the whole project?
	FRANZ FISCHNALLER: Among the most interesting challenges encountered over the course of this project, the City Cluster design calls for a VR environment that allows users to be in two different cities at the same time. Meta-Net-Page is a virtual-reality networking interface display, designed and implemented ad hoc for City Cluster. This "view panel" will serve as the main interactivity tool for the user. Is a virtual pathfinder, able to detect information, images, and details that are invisible zones or intangible realities for the naked eye. It indicates the current user coordinates and provides more detailed information about the objects within its view. It will allow the user to zoom in or out by moving the panel closer or farther away from their own eye. The user will be able to fly up into the sky to gain a greater overview of the city they are currently exploring. Then, using a button on the side of the Meta-Net-Page, the user will then be able to "teleport" immediately to the location shown on the view panel. The Meta-Net-Page allows multiple viewpoints to be rendered of geometry that may or may not be visible. The viewpoint or the user's panel, the viewpoint of his/her net partner's panel, and a third external viewpoint, seen by the user themselves, can all be represented at once. Conceptions of "local" and "remote" begin to dissolve as visitors share multiple perspectives through the Meta-Net-Page. MARIALUISA PALUMBO: About the interactive platform, the application is designed to run in the CAVE (Cave Automatic Virtual Environment) and on the AGAVE display system (Access Grid Augmented Virtual Environment). Can you tell me something about the main features and differences between the two systems? FRANZ FISCHNALLER: The application can run in both platforms because the Agave is a derivant of the Cave. Same hardware and same software and technological capabilities. Except that the Cave has for wall porjections and the Agave has one wall projection. The CAVE (CAVE Automatic Virtual Environment) is a projection-based virtual reality system, it is a 10 foot-cubed room. Stereoscopic images are rear-projected onto the walls creating the illusion that 3D objects exist with the user in the room. The user wears liquid crystal shutter glasses to resolve the stereoscopic imagery. An electromagnetic tracking sensor attached to the glasses allows the CAVE system to determine the location and orientation of the user's head. This information is used by the Silicon Graphics Onyx that drives the CAVE to render the imagery from the user's point of view. The user can physically walk around an object that appears to exist in 3D in the middle of the CAVE. The user holds a wand which is also tracked and has a joystick and three buttons for interaction with the virtual environment. The buttons can be used to change modes, or bring up menus in the CAVE, or to grab virtual objects. VR applications displayed on the CAVE can be linked over high-speed networks. The AGAVE Access Grid Augmented Virtual Environment display system is a passive stereographics projection system through which Audiences will view the immersive content using 3D movie glasses. The display system consists of a pull-down polarization-preserving silver screen, two LCD projectors with linear or circular polarizers in front of each lens, a dual processor Linux driven PC with a high-end graphics card capable of dual display output. To support 3D tracking and interaction, Ascension Technology's Flock of Birds (Extended Range Transmitter version), and EVL's Wanda can be used. Wanda is a wand with an imbedded tracker, joystick and three buttons. It is the standard interaction device used in CAVEs. The overall concept behind AGAVE is to append PC-based graphics workstation to an Access Grid node that can be used to project 3D stereoscopic computer graphics to allow networked people to immersively share three-dimensional content. VR applications displayed on the AGAVE can be linked over high-speed networks. Both platforms interconnect and run over high-speed networks, enabling local and remote the visitors to navigate, interact, and communicate with each other in real time through the avatars in a common virtual space. Passive and active stereoscopic glasses allow them to see in 3 dimensions. A simple tracked input device, "the wand," containing a joystick and buttons, allows the visitor to navigate through the VR space, to manipulate virtual objects and to interact with one another.

	> CITYCLUSTER

	interview