Creating Avatars

An avatar is a representation of the viewer in the virtual environment. In Ygdrasil an avatar is called a "user", and is encompasses a set of nodes which define both your visual appearance and the way in which you navigate and interact with the world. The user nodes allow you to load an object for your head, your body, and your hand; turn gravity on and off; and control how fast you can move, among other things.

The user is constructed through several nodes grouped together:

The visual representation of the avatar is divided up into the body, the head, and the hand. This organization comes from the usual configuration of the CAVE - one sensor tracking the movement of the viewer's head, and one sensor tracking the movement of the viewer's hand. This means that the avatar, the virtual body, moves in exact response to the user's movements. Wave your hand around, while holding the wand, and the avatar's hand moves too; turn your head, and the avatar's head turns too. This visual representation is a crucial part of the perception of presence, or immersion. It allows the sensation or experience of embodiment in the virtual space, particularly to the degree that the interface is literal rather than metaphorical.

The interface is of course not completely literal - the virtual body is not a one-to-one map of the real body, and it's interesting to consider the weight and importance placed upon the chosen interface points. In this configuration, the 'body' is reduced to two points, the eye and the hand; everything else vanishes.

In Ygdrasil, the "Body" node is in fact not even responsive to the sensor interface as the head and the hand are - it's just an object assumed to be a little bit below the head. Other body parts are similarly ignored. An "arm" is often constructed by guessing at a point for the shoulder, based on the position of the head, and drawing an object between that and the known position of the hand; "legs", on the other hand, have almost no sensor data to work with.

Of course, how much time to you spend looking at your own elbows and knees? How often do you look at the back of your head? In a virtual environment, without programming in mirrored surfaces, you can never see your head at all. However, there is another reason to create a visual avatar - it is a representation of yourself for other participants in a multi-user world. Ygdrasil is designed from the ground up for networked, collaborative, multiuser environments, so the avatar is additionally important for these types of applications.

Ygdrasil, and the CAVE in general, also differ from the conventions of many familiar virtual 3D graphics environments in adhering to a strict first-person point of view. Many games, for instance, use a third-person or "over the shoulder" point of view, where the camera flies behind and often slightly above the player's actual body. In these scenarios, you do in fact spend most of the time looking at the back of your own head. First-person shooters don't follow this convention, but many still include the option of a 3rd person point-of-view. There a number of reasons for choosing one or the other type of camera viewpoint, but the CAVE and other virtual reality systems tend towards the first-person approach since they are predicated on a direct transference of the physical body into virtual space to create the sensation of embodiment.


(c) Ben Chang