Light Painting The Goodby Silverstein Office

As a summer intern project, we wanted to do something badass. So we decided to create a light painting of the Goodby Silverstein & Partners logo across the entire front of our office building.

Light painting is the process of generating images by moving light sources across a long exposure photograph. It’s been used for some very cool ads over the last few years.

To generate an image with light painting, you can program a strip of LEDs to flash one column (or row) of pixels at a time. If you move the strip at a constant speed, you can reproduce an image in the camera exposure.

We decided to hang a long column of LEDs from the roof of our building, and move them horizontally, gradually revealing the logo in a long exposure.

The LED strip is 8 meters long. It’s made of 256 individual LEDs, all of which are controlled by a Raspberry Pi B+. Both the RPi and the LEDs are powered by a single 5V/10A power supply, which runs up to an outlet conveniently located on the roof.

(left) The Raspberry Pi and wiring to the top of the LED strip. (right) All 8 meters of the LED strip laid out.

Using a script 100% borrowed from Adafruit, the Raspberry Pi can parse a PNG image file, one column of pixels at a time. It then sends those pixel color values to the LEDs at a rate of your choosing. We used a 20 second exposure time, so the full image took about 18 seconds to cycle through.

Dangling the LED strip off the edge of the roof.

We had a team hang the LED strip from the roof and walk it horizontally across the exposure. Two cameras on the ground captured it from different angles.

In real time it doesn’t look like much.

But taken with a long exposure…

The hardest part is timing. The LED strip has to be moved at a perfectly constant pace. Because of wind and edges on the building, that can be hard to do.

We had lots of outtakes.

But in the end, we got it! (You can see us in the top left, above the “G”)

The GS&P Interns are Elliott Spelman, Jasper Yu, Candace Faircloth, Scott Menzie, Emily Ng, and Melanie Matlock. Special thanks to Issac Le, Quinn Gravier, Patrick Wong, Celine Chappert, and Leslie Bee.