> Motivation
> Example Movie
> Comment
> Conclusion
> Related Experiments
> Setup

Performance Photo


> Top



In cubism different fragmented perspectives allow us to piece together an image that is not restricted to the view from one direction, but as if viewing it simultaneously from all sides and thus getting more complete information on the depicted object. We asked ourselves if this principle applied in cubism would also work for the movement of a dancer when constructing the body out of images from different points of view. A limb that - from the spectator's point of view - is invisible behind the dancer's body, can still be visible in the projection.


> Top


Example Movie

Stage View
Computer Image
A short choreographed duet was captured simultaneously by two cameras positioned in the front and on the left side of the space. The front-view of one camera is fused with the side-view of the second camera, depending on where the sensor (on Nik's hand) is in space.


> Top



In a simpler version this setup could have also worked by superimposing the images of the two camera angles. But in this experiment it is the dancer's movement that determines if we see and where we see one or the other viewpoint.

For us the second half of the video excerpt (after the first 30 seconds) was the most interesting to see. Here the two viewpoints from front and side meld and form an odd mix of body positions and movement. Changing between moving and remaining still in a position develops an interesting dynamic.

What is revealed here are two viewpoints of the dancers during the choreography. They usually appear alone, or sometimes one and some fragments, which allows their world to remain enigmatic. This shared space, when looked at in the right side of the screen, is as real as the physical duet but which does not show up in the left screen at all.


> Top



The superimposition of two viewpoints conflicts with our visual perception and has to be deciphered. A static or slowly moving dancer in the center/ intersection of the camera views gives the recipients enough time to decipher the construction. In this case really a spatial quality is added to the representation of the dancer, even though the construction perceptually stays abstract. Fast motions are to difficult to decipher and are becoming annoying very quickly.

Since the superimposition only works if the dancer is located at the intersection point of both camera views, action is limited to a very close area, in our case a fixed location on stage.


> Top


Related Experiments:


> Top





  • Graphics PC with 2 Video Capture Cards
  • Projector
  • Polhemus tracking system: 1 Sensor
  • 2 Video Cameras


  • One Video Camera is facing the dancers. The second camera looks at the stage from the left side. Both cameras transmit their live video image to the graphics computer. The graphics computer displays the image from the front camera on the projection screen. This image is overlaid with the image of the left camera, which is shown within a round spot. As illustrated by the Technical Screenshot below, both images are superimposed within the area between an inner and an outer circle. The spot behaves like a smooth mask that shows the front video image in the outside and the left video image in the inside.
  • The Sensor controls the position of the spot on screen. This is how the dancers are in control of which fraction of their bodies are displayed from the front view and which from the left view. In the recording we attached the sensor to the right hand of one of the dancers.


Technical Setup

Spatial Setup

Technical Screenshot

Performance Photo
> Top