Dancer Position and Marcov Chains

by
Will Wolcott and
Anne-Marie S Hansen

This project is intended as a scenography setup with 1 to 3 dancers on the scene.

The visual output is organized by Brownian motion, where object positions are randomly assigned, but are based on the position of its neighbor. In addition, the objects begin a random walk, causing them to tend to scatter in the image. The constant Brownian motion of the system leads to an unatural and undettermined behavior. This happens in parallel to the disorganizedd nature of a camera tracking dancers' movements. As dancers begin to organize and repeat behaivors, the visual system responds in kind. The objects in the image stop their persistant movement and settle to the neighborhood of the entire system.

1) When the dancers move on the scene, a camera will track the dancer or the middle position of 2 to 3 dancers on the scene and animate the graphic shapes towards that position on the projection.

2) When the dancer is close to the camera, the shapes will move towards the viewer - they move away from the viewer, if the dancer is moving away from the camera.

3) The camera will recognize if the dancers are standing in a position and repeat that position. When a position is recognized one time, the shapes will start moving less randomly. When that same position is recognized a second time, the 'movement noise' or 'shape noise' will be totally gone.

The same position is recognized once, and there is less visual noise on the picture.

The same position is recognized twice, and the shapes have no visual noise left. When the dancers stand still the circles move into a large formation on the projection.

The first pictures shows dancer position far away from the camera - the second shows dancer position close to the camera.

The dancers are close to the camera, while they move from the side into the center of the scene.

The dancers moving further away from the camera from the center into the side of the scene.