Chronopticon: A look back in time

For an exhibition at Peron E in Utrecht, Romy Koch and I were asked to create an installation with the theme “white.” After a months-long process of brainstorming, researching, prototyping and refining, we created the Chronopticon. It’s a machine that allows viewers to examine how the present—no matter how ordinary it may seem—is the culmination of a forever-long, colorful past.

The concept

Romy and I were given the theme “white” and asked to explore it. We started by making long lists of every association we had with the word, from objects to feelings to cultural implications. Through many of these entries, we realized that white was often quite contradictory. We saw it as a color of peace or surrender (a white dove or white flag), or blankness (a house with white walls), or dullness (ambient white noise).

However, beneath these association is a process of chaos or accumulation. The white flag of surrender waves after battle. White, as we see it on our phones and TVs, is a product of the other visible colors. White noise can be broken down into energy across a spectrum.

And we applied this to the spaces we inhabit every day, too. Even alone in our homes, we view every room and object as this moment’s representation of a long, storied past.

With this in mind, we wanted to create a space where people could explore what had happened there and understand the broad spectrum of events prior to their arrival.

The installation

To present Chronopticon, we set up a large, white tent in the exhibition hall. Inside, we placed a remotely controlled camera and a custom-made interactive display built around a Raspberry Pi and Tascam DR-40 field recorder.

When we switched it on, the camera would take a photo of the empty interior. This photo would serve as a reference image for the Python script that ran the installation. As people would wander in and out of the space, the camera would snap a photo every few seconds. Whenever a new presence or a person’s movement was detected, the program would find the changes in the individual color channels of the image, and it would turn the pixels with changed values to 255—the highest in RGB color. These changed images were then layered over each other. A similar process took place with audio clips recorded with every photo.

People could view the images at the display, winding backward and forward to see how the layers accumulated. Over time, however, the image would become whiter and whiter as the photo masks piled up. Additionally, the sound would become closer and closer to a wall of white noise.

The tech

The installation was controlled with two Rasbperry Pi computers: one inside the Chronopticon console and another connected to the camera on the opposite side of the room.

Using gPhoto2, we were able to control the camera via USB with a simple shell script called at the start of a Python loop. Once an image was snapped and downloaded to the Raspberry Pi, Numpy was used to compare the RGB and Alpha channels as arrays to the background image. Once color changes outside an adjustable threshold were found, the image channels were shifted towards white, and the corresponding pixels were turned into an image mask using the Pillow library for Python.

This mask was added to the most recent image (which was already a collection of all previous masks + the background … ) was saved down into a new photo and sent via a simple server to the client machine inside the Chronopticon.

The interface allowed visitors to use a large wheel connected to a rotary encoder below to cycle through the images and sounds that had been collected. Within about a minute, people would also see and hear themselves added to the room’s long history.

For more images, you can view Chronopticon on Instagram: @chronopticon.