I began working on "The World Generator / The Engine of Desire" five years
ago - before VRML. I had a sense of where virtual space was going and
wanted to articulate it. I often write texts in this kind of "list" form. I
worked with Gideon May, an exceptional programmer - presenting him with an
elaborate plan for a new interface. We were supported by the Australian
Film Commission. I was living in Australia for many years. Australia was a
fantastic environment for this kind of work - many many interesting and
talented technological artists... anyway - now the work exists. A new
version of the "The World Generator / The Engine of Desire" will open with
the ZKM in Karlsruhe - Oct 18th.
I am currently doing a Ph.D. at CAiiA - TheCentre For Advanced Inquiry Into
Interactive Art at the University of Wales with Roy Ascott. It is quite an
exciting program. My title is --
"Recombinant Poetics: Navigation and Poetic Construction in Interactive
Art." Here is a technical description of the work which houses the text.
The World Generator / The Engine of Desire c96/97
Bill Seaman , Objects, Video, Texts, Texture Maps, Music
Programming by Gideon May
Sax and Clarinet by Tony Wheeler
Initial Investment through the Australian Film Commission
"The World Generator / The Engine of Desire" is an interactive
computer-mediated environment which enables viewers to construct and
navigate poetic worlds in real time based on an interactive template of
potential choices. The system is facilitated through a new interface
metaphor. At the bottom of the screen is a rotating set of
container-wheels. These container-wheels house a variety of selections. The
menu system contains the following set of wheels: 3d models; poetic text
fragments; texture maps - both still and video; location sensitive audio
objects(musical loops); behaviours; and function menus which enable the
viewer to center themselves in the world; to scale objects and texture
maps; to make objects and texture maps transparent; to construct [semi]
random worlds; to make a series of different "random" choices including
random text, random sound, random object, random texture map, random movie;
as well as to clear the world. The still and moving texture maps can either
be applied to objects with an "aura" or be placed in the space as pictures
and/or movies on flat screens.
The variables carry specific artistic content or fields of potential
meaning. After generating a world, the viewer can then enter into the world
constructed from their selections, and navigate as well as manipulate the
chosen variables within 3D space. This suggests a new spatial cinematic
form. At present only one video texture map plays at a time (based on
viewer proximity). When the viewer moves away from the video, the selection
becomes a still. As the viewer approaches another, it becomes active.
The data projector displays a "World" which is manipulated from this
construction menu at the bottom of the screen. Objects which are selected
from the construction menu are instantly entered into the "World" space in
real time, directly in front of the viewer. A transparent "aura" defines
the activated object (the object the viewer can manipulate/ edit). the
viewer can toggle the "aura" on and off from the space mouse. The viewer
can also move close to an object and select it by toggling on the aura.
They can then act upon their choice. The viewer is able to control the
rotation of the menu wheels, moving them forward or backward, accessing
many choices with great ease.
Each of these rotating menus has a particular aesthetic. Thus the set of
rotating menu wheels becomes a rotating icon series. A specific number of
elements (depending on the quality of the computer) are present in the
space at one time. When the "Random World" choice is made, a world is
instantly constructed with 25 objects (the current setting), including
random texture maps and sounds attached to them. A selection of Random
Behaviours applies a random behaviour to each of these objects. If the
viewer keeps hitting the random button, old objects are deleted as new
objects enter the space.
Process
The work functions in a two stage process: The viewer first constructs a
"poetic" environment based on selections from the template of variables.
One mode allows the viewer to observe a blow-up of the menu. When the
viewer chooses objects from the main menu, a set of container-wheels
containing pre-rendered 3 D objects rotates. The actual storage is in the
form of long "virtual" rotating belts which can have great length (based
on available memory) although the viewer only sees the curved front end of
the belt. The following manipulations can be made: Layout - placement of
the object on the site; Scale- scale of object on the site; Texture Map -
attach still or video texture map to an object with an aura, as well as
place "screens" of stills and video into the space, when the aura is
toggled off; Behaviours - attach behaviours to selected objects, texture
map screens, and/or sounds. Once this process has been completed (or
anytime during the construction process) the viewer can enter the space and
navigate. An optimised version will be made for WWW use with the potential
of attaching WWW addresses to the objects making it a functioning 3D
browser.
An elaborate object based text is included in The World Generator. A viewer
can choose any single line from the text and place it in the space as a
visual object with a "location sensitive" audio text triggering mechanism.
I have entered my voice speaking each line of text. As the viewer moves
close to the 3D text they hear the line spoken. Thus a new spatial literary
form is being developed. Texture maps and behaviours can also be applied to
these text objects.
Recombinant Music - The sounds included in the system are made up of
hundreds of techno ambient loops, composed by the author, consisting of
synthetic rhythms, drones, and tonal loops. Specific tonal sax loops have
been played by Tony Wheeler. These sound object loops are placed by the
viewer on the site. A circular "drop" motif allows the viewer to see where
the sound has been placed (if it has not been attached to an object). The
visualisation of these sound positions can be turned on or off through the
"show sound" / "hide sound," menu choice. As the viewer navigates, a
location sensitive audio mix is generated. As one moves close to a sound
object the volume goes up, as one moves away the volume goes down. As the
viewer navigates by an object on the left - the sound comes from the left.
If a viewer navigates by a sound on the right, the sound comes from the
right, an object in front of the viewer sends sound to both ears equally,
making a stereo mix. A behaviour can be attributed to a sound object. A
sound object can also be attached to a 3D object.
The work can be interacted with from multiple locations. Two or more users
can be involved with poetic construction and/or navigation in the space at
the same time, currently via modem, and in the near future via the WWW. I
have coined the term "RE-I" (pronounced RAY or Re --I) short for
Re-embodied Intelligence, as the term to describe the visual representation
of these alternate, multiple users. A "RE-I" is potentially visible in the
space, showing the location of the alternate user.
Bill Seaman
-------------------------------------------------------------
a forum on spatial articulations, perspectives, and procedures
texts are the property of individual authors
for information, email majordomo@forum.documenta.de with
the following line in the message body: info blast
archive at http://www.documenta.de/english/blasta.htm
or http://www.documenta.de/deutsch/blasta.htm
documenta X Kassel and http://www.documenta.de 1997
-------------------------------------------------------------