Ananke
An Interactive Dance Performance Using Infrared Light Tracking made in openFrameworks
Made in collaboration with Steven Wyks and Erica Wu
Ananke, a reference to the greek personification of inevitability, compulsion, and necessity, deals with the double-edged nature of technology, in the midst of a society where the lines between technology, surveillance, and data privacy get increasingly blurred. The piece deals with the interactions between a dancer and a mysterious dodecahedron hanging from the stage ceiling, triggering sounds and visuals throughout the performance. The performance starts serene and playful as the dancer gets used to the seemingly innocent dodecahedron (which represents technology and social media). However, as she invests more time and focus on it, a hidden figure emerges, suggesting that what initially seem harmless might not actually be so.
Technical Specifications
The basis for the interaction is the live tracking of infrared LEDs inside the dodecahedron using a PS3Eye Camera. The dodecahedron is lit up by a Neopixel ring attached inside. The Neopixel ring changes color manually via a Bluetooth LE and the Adafruit mobile application. Based on the location of the dodecahedron, various sounds are played and eye-catching visuals are projected onto the wall.
Technical components:
Laser-cut dodecahedron - laser-cut and fabricated by ourselves
PS3Eye Camera
IR LEDs
Teensy LC
2 Lipo batteries
2 PowerBoost 1000 chargers
Neopixel Ring
Bluefruit LE + Adafruit mobile application
openFrameworks - used particle systems, attractors, brightest point tracking, thresholding, ofxCv, and blob detection
Visuals
All the visuals done in the performance were done by scratch in openFrameworks. Particle systems and attractors were used for the foundations of the visuals, as they provided the most flexibility and consistency throughout the different stages of the performance.
Tracking
We tracked the dodecahedron by converting the image received from the PS3Eye Camera to a grayscale one, filtering out the background and making the brightest points more prominent in the image. This was then passed through a contour finder using ofxCV, and the x and y location of the first contour in the array of found contours. This is then mapped from the size of the camera’s dimensions (640 by 480) to the full width and height of the window. We added a debug mode which would allow us to see the camera as well as a mode that would allow us to control the graphics through the mouse rather than the tracked dodecahedron.
