Construction
Construct: PUSH
Previous Chapter | Next Chapter
AimPush is a prototype of a possible way to interact with an embedded space. It is a critique of The Virtual Architect and the digital virtualization media at CAVI (Center for Advanced Visualization and Interaction).
Since The Virtual Architect was made as a video prototype by the use of chroma key technique to layer two sources of video, the actual level of interaction was very low. We only saw the resulting embedded space on video monitors in the studio, why the interaction were played out with one eye on the monitor and one on the blue screen. With Push, we wanted to go one step further and construct a direct interaction with the data and if possible use it to design. We wanted to use the same software as in The Virtual Architect, but in the Panorama of CAVI – a 180° visualization screen for constructed spaces.
First, we wanted to select the parameters and operands and to define the degrees of freedom, which together with the rules of interaction, would structure the embedded space. Instead of focusing on the wide range of data as in The Virtual Architect, we wanted to focus on simple interaction with very few data objects. We also wanted to use the Panorama in a more creative and constructive way, than just to visualize how the data objects would be placed in space.
Second, we wanted to question the use of space and the creation of space through manipulation of objects, with a reference to the 19th century architect Gottfried Semper. He argued that there was two ways to construct space. In a tectonic way, the architect gathers objects and positions them in a specific order to overcome gravity and to create enclosure. In a stereotomic, way the architect removes or repositions objects that are already there to create space. These two ways of construction can be found in the very fundamental erection of a structure in architecture and in the caving of spaces in dirt or stone. Architecture is today primarily a tectonic construction. However, in a virtual design environment it could just as well be a stereotomic creation of space, as in the concept of ‘data mining’ as described by Sara Fabrikant.
In this way, the prototype became a critique of the techniques of the media we had chosen for our construction. We could compare this to the technical possibilities that defined Robert Barker’s panorama and to how the interaction between media and user could be constructed, within the constraints that the Panorama at CAVI offered.
ConstructionBefore we started it was crucial to know the constraints of the medium that we used. In this case, it was a panorama with three projectors with interaction tools mapped in space by ultrasound and driven by the same SGI farm as in the chroma key studio. The amount of data was limited to 50000 polygons at 60 frames per second. We were limited to simple triangular geometry of both space and objects, but could import the same data objects that we had used in The Virtual Architect. There was a difference between the constraints for manipulation of viewpoint and manipulation of objects placed in the space. The view point has 9 degrees of freedom (x,y,z in position, rotation and scale) while the objects only have 6 degrees of freedom (x,y,z in position and rotation).
Unfortunately, we were never able to test the ideas in the actual environment, so Push was kept as a proof of concept and a short video prototype. The main bottleneck was the specific programs and scripts that would control the interaction between the data in the embedded space and the real environment of the Panorama. Therefore, all illustrations were constructed from a parametric study model in MAYA, which was used as a virtual laboratory with the same constraints of movement, rotation and scale as in the Panorama.
DiscussionWe learned from our experience with The Virtual Architect that we quickly could reach an alarming level of semantic overload if the data objects were too dense. In a virtual and stereotomic space, this would lead to the elimination and ordering of the data that was already present. This is a phenomenon that Sara Fabrikant has described as “information pollution”.(1) The best result in The Virtual Architect was to limit the data sets to two or three layers, which were easy to distinguish: One for a position reference, one for textural notations and one for designing or investigation. In Push, location in space was crucial for the initial position: Not an absolute position, but rather a relative position to a recognizable data object, which could be given a specific color, label or texture.
Another issue was that almost everything was possible in the chroma key studio as it was a layered video production, while very little was possible in the panorama studio as real interaction with the data objects. In Push, we were therefore not able to reach the same easy design flow with Nurbs and skins as we proposed in The Virtual Architect: We had to produce a new way of designing, hence the focus on the difference between stereotomic and tectonic spaces.
The video prototype is a joint experiment between two overlapping PhD. projects (Thomas Leerberg and Tina Henriette Kristiansen), which both deal with the architectural consequences and possibilities of a computer embedded reality – both immersive and augmented.
On the DVD with video and image material in the back of this dissertation, there is a short video that shows the ‘chain of boxes’, which was one of the ideas for the construction.
Previous Chapter | Next Chapter
(1) Fabrikant, Sara Irina & Barbara P. Butterfield (2001), 'Formalizing Semantic Spaces for Information Access', in Annals of the Association of American Geographers, 2001/91, Malden, MA: Blackwell Publishers. p. 264.
|
|
