Carnival Glass

Carnival Glass Thesis

Inventory list: webcam, computer, microphone, screen,
Programming: Wekinator and Max Jitter programs
Medium: Video
Dimensions: 90x60cm

Carnival Glass Proposal
Carnival Glass Image


Interactive • Machine Learning • AI

This portable sound sculpture is affected by the temporal nature of the city and its occupants. The Carnival Glass fun-house mirror aims to visualize the real-time ambient sound of a location. Any verbal interaction with the system should cause spectators to reflect on the issue of noise pollution.

Carnival Glass captures ambient sound with a microphone which motivates video distortions on a screen. Max Jitter works together with machine learning to understand the tone and intensity of the sounds. Predetermined parameters in Max uses data from Wekinator to distort or sharpen the image. As people interact with the mirror, the sound of their voices overcome the distortions clearing the screen and smoothing the image.

Like any fun-house mirror, it is up to the spectators to determine why the distortion is happening.


Scroll to Top