Things I’ve Learned

Musical group OK Go released the video for their song “Here It Goes Again” in 2005. The iconic video shows the four band members dancing wonderfully on eight moving treadmills. Shot with a single locked-off camera, and now exceeding 50 million views on YouTube, that video redefined what a viral video could be.

In August 2009, Syyn Labs began discussions with the band to build them a machine they could “dance with” in a Rube Goldberg-style chain reaction for their next video. A few requirements: no “magic,” and the machine should try to hit beats throughout, play part of the song, and be built to be photographed in one continuous shot.

The build was daunting, but ultimately it was a great success. (See the video at bit.ly/okgosyyn.) Here are some things we learned.

1. Do the small stuff first. We wanted the machine to build in excitement, to crescendo with lots of big, crazy interactions. Also, we’ve found that …

2. Bigger is better. Smaller components are more fidgety than larger objects. A marble and its trigger are simply far more affected by dirt, temperature changes, and vibration than a bowling ball, which doesn’t much care at this scale. Therefore we also …

3. Put the less reliable stuff up front. This was important, so that we spent as little of the precious shooting time resetting the machine as possible. With 89 different types of interaction, and many times more than that if you count each physical interaction (each domino, chair, rat-trap flag, etc.), we wanted any failures to happen near the beginning. Nevertheless, it’s important to …

4. Have lots of people involved. Ultimately, we had more than 55! They were all essential and worked long hours late into the night to get everything working beautifully. Of course, when you have that many people working on a machine so large, you must …

5. Assign teams. This machine was really big, so having a dedicated team for key components helped improve reliability and minimize danger (from falling pianos and steel drums). This specialization also allowed for flexibility when last-minute changes were needed. Why the changes? Well …

6. Aesthetics are important. Some interactions in the machine were too fast, or too crowded, or just didn’t have the right “feel” on camera. For example, the piano was intended to come down slowly, but it looked so good crashing down that we decided to make that change. Unintended consequences occured due to vibration, but a workaround was discovered, and we continued, proving my last point …

7. Be flexible. Early on, we knew we wanted to end with paint cannons, and we suspected we’d start with dominoes (a classic!), but the rest was a realm of infinite possibilities. Many ideas were pursued and abandoned: some were hard to photograph, others simply didn’t work. It’s hard to give up on an idea after days or even weeks of work are invested, but, we found, it’s often necessary for the greater good.

More photos at: makezine.com/24/learned