Parameters for Crossings

Chiron's programme is the basis for the first workshop. The programme defines the behaviour of a set of agents moving from one side of the space to the other. The programme is interfaced with a camera interface allowing the agents to see brightness values in the real world. When the agents come across dark patches (moving or not) they will avoid these.

The agents move in respect to vision: the agents see the goal (attractor) and move towards it. In Crossings the agents see the opposite side of the space from which they are born. Moving across the space they avoid to bump into each other. This creates neat paths negotiating the space. In Crossings this is visualised at a trace. At present this is as a literal trail, in the future this could be as a particle stream emitted from the agent itself.

Behaviour of the agents (no trace and no tracking)

Behaviour of the agents with trace (no tracking)

Behaviour of the agents with trace and tracking

The parameters for the agents are:

Space:

The agents have a certain size and movement quality by which they seek the attractor. This is controlled mostly by their number (e.g. how many they are) but also by the routes and distance they have between them.

The space of the agents will be highly determined by the choice of visualisation. At present we are thinking this as a particle stream affected mutually by the behaviours of the agent and the passerbys.

At present the parameters of the agent space is:

Agent number: e.g. how many agents exist in the space
At the moment we have quite a lot of agents (50 -200) to fill the space but as we change the visualisation this might be severely reduced (e.g. 10 Ð 15).
Agent proximity: how close is too close (which parameter sets this Ð if at all?)
Avoid common routes: makes the agents always seek new routes.
Turning distance: this controls the movement quality of the agents Ð e.g. how rounded they move.
Pulse variable: in Crossings the agents move in pulses emulating the idea of the zebra crossing. The variable sets the time which the agents are given to find their attractor. If the agents are too slow they are reset at the end of the pulse. (wait for all has to be ticked)
Boundary: the space of the agents is set up within a boundary which ensures that they don't move out into endless space.
Question: I cant really find a way of setting their movement speed Ð is this just me?

Trace/particle stream

At present the agents leave a trail behind them as they move in space. We are intending to use the agent as a driver for a set of particle streams generating a flow of movement. The particles streams are sited in an environment which holds certain properties such as gravity and perhaps vortexes (these might be better conceived as active attractors perhaps mapped (in some instances) to the presence of the performers).

Following ChironÕs suggestion the particle system will be determined through its flow across an (invisible) surface. The surface is informed by the interface as well as the agents behaviour (speed of movement?). As the movement across the (Mixed Reality) surface slows down the surface deforms creating deep crevices. The surface is continually in a process of deformation thereby controlling the movement of the particles.

Vision

The agents operate through their ability to ÔseeÕ their environment. This ability is defined by the following parameters:

Initial vision distance: what the agent sees at first, if at 1 this means that the agents see nothing and are therefore paralysed Ð e.g. they are not attracted.
Variable vision distance: not sure?
Follow agents: means that the agents follow other agents that might be able to see rather than relying on their own vision. This creates local spaces or neighbourhoods, meaning that one agent might lead another astray.
This is obviously really nice Ð but without Follow Agents and with a high Initial Vision Distance what is also nice is the sense of them being lost Ð not seeing that which is right in front of them and instead moving towards a further goal.
Meandering and meandering degree: what does this really mean?

Interface

The agents move on a white surface which can be mapped to the input from a camera. In the this case the changes to this white surface (e.g. people moving) will be tracked as boundaries, meaning that the agents will avoid them. (we can set the degree to which the agents will avoid these).

Detail of trace: notice the way the traces already form little eddies.

The camera interface has the following parameters:

Size of video (should be at max e.g. 640 *480 and the space of the agents should be directly mapped to it.
The whiteness of the background image can be set within the application (light threshold) but also with the cameraÕs own settings (e.g. Logitech programme).
We can set a difference image for this so that we are operating with the subtraction of new events rather than old events.
We can choose to see the video or not.

Phase shift: creating performance sets

The idea of phase shift and [1] is the change between how entities perform in respect to a set of conditions, moving from one order to the next. This might be a helpful term for us as learning from The Changing Room we will need to introduce the idea of different states into the performance. In The Changing Room this was made through scenographic, choreographic and digital changes (e.g. morphology and colour) to the performance. In Crossings a phase shift could be induced through a change to the way that agents perceive the world, e.g. what do they find attractive and how do they navigate/see the world.

Similarly we could introduce the idea of bifurcations (a splitting of how the system reacts into two simultaneous reactions).

Setup for workshop

[1] A common example is the gas-liquid phase shift undergone by water. In such a transition, a plot of density versus temperature shows a distinct discontinuity at the critical temperature marking the transition point. Similar behavior can be seen in systems described by ordinary differential flows and discrete mappings. In nonlinear dynamical systems, the transition from self-organizing to chaotic behavior is sometimes referred to as a phase transition (or, more specifically, as an order-disorder transition).