FIT5900 Lecture: Virtual Worlds

Being There

Virtual Reality (VR) aims to provide this experience synthetically by fooling a viewer into experiencing a space where there is none.
Often digitally synthesized, animated images are presented to a VR inhabitant but...

Early text and text / still-image based adventure games like Zork, Beyond Zork, The Hobbit & Transalvania were extraordinarily adept at transporting the player to another world for real-world hour after hour.
A still image creates a kind of virtual reality and may be used to give the illusion of space.

To make a virtual world inhabitable consistency is vital!
- Consistent behaviour (objects / characters)
- Consistent location (for navigation)
- Consistent interaction
An inconsistent world is impossible to inhabit comfortably!
It is also beneficial if there is a purpose to one's visit to a virtual world.
- Explore (only if the world warrants it!)
- Solve a puzzle (only if the puzzle warrants it!)
- Chat with other inhabitants (only if the inhabitants warrant it!)
Provide sophistication of interaction...
- Via a story tied in with a puzzle... hence the success of text based adventure games!
- Complex and rewarding interaction through the inclusion of other human inhabitants... hence the success of Doom, Quake, Internet chat rooms etc.
- Complex artificial agents are also becoming viable inhabitants for interesting worlds. (Space Invaders & Pacman ghosts were just the beginning!)

The domain of Computer Graphics concerns itself with Image Synthesis... the modelling of illusory light reflected, transmitted and absorbed by objects which themselves do not exist, like these cubes...
Most 3D graphics is produced using off-the-shelf software with a GUI that even a newcomer to computers can decipher after a few years of training, headaches, heart aches, lost clumps of greying hair, broken relationships etc.
3D software packages may provide means for:
- Modelling - the process of specifying the geometric properties of an object.
  
  Usually performed through a GUI allowing the user to select basic primitives (cones, spheres etc.) and modify their geometry by various means as well as decorate them with textures and connect them in hierarchies.
  
  At this stage synthetic lights are also added to the model for the rendering process.
- Animating - the process of specifying the time varying properties of a model.
  
  Usually performed by a keyframing process where the model is posed at various times and the computer mathematically determines where the model ought to be positioned between these key frames.
- Rendering - the process of synthesizing images of a model.
  
  Usually a computationally expensive process where imaginary light rays are bounced off the imaginary model at each time an image is required and for each pixel in the image to determined the colour visible to a viewer at that point in space.

A means of illustrating a world (real or imaginary) using a panaoramic photograph or rendering 'stitched' into a cylinder. (Other similar methods stitch images into complete spheres)
Not a computationally expensive thing to display in a browser!
Users can interact with the pseudo three dimensional world using the mouse.
Objects in the world can't really be animated.
Locations on the image may be hotlinked to other QTVR scenes.