Pin Ball Gizzard

I have a project in mind that makes use of a recently reclaimed pinball machine (thank you streets of New York). Said pinball machine was thrown out with other interesting detritus from the Ricky’s on 84^th St. and Columbus just as the school year was getting under way. Thankfully, my well muscled and good tempered friend Andy lives just down the street. The coolness of the broken pinball machine was enough to convince him that this thing needed desperately to live in his apartment – and there was light.

This is not Andy.

I’ve been thinking about the busted pinball machine and have come up with a fun final project idea for ICM. It would have two user states; first, as a pinball machine with added graphics and monster sound and second as a music player (for Pandora, perhaps) while the machine was not being played. For the midterm, however, I have decided to attack one of the graphic ideas I had for the screen I will be installing in the pinball machine.

It is inspired by the gloriously beautiful video “WTF” by internet video gods OK Go, which I later learned was made with…. PROCESSING!!  They made a video using a single take of a live interactive video echo of sorts. The foreground where the performers moved was entirely visible, but the background was comprised entirely of previous frames showing what had last occupied that space of the composition.

Knowing that this wasn’t nearly as easy as it looked, I was interested to see if I could create an effect that was similar in construct for the graphics on the pinball machine. In my video, however, instead of keeping the foreground images in their original colors, the silhouettes would be opaqued out with a single color. The background would also be a solid color. Another difference is that in my video, I wouldn’t be using a green screen and would need to cancel out the background.

To do this, I decided that I first needed to take a snapshot of the background where the player would be standing (without the player, of course). Then compare that frame to the current frame. If the pixels were the same, then the pixel would be colored as a background pixel. If they were different, then the pixel would be colored as a foreground pixel, giving us a monochrome silhouette of the player in action. The next time the loop ran, the different pixels would lay over the old silhouette in a different color. Hopefully giving a multi-colored echo to the composition.

However, despite the hours I have spent banging my face against my keyboard, I have yet to achieve the desired effect. Having seen what I have in my experiments, I doubt that I will be able to create precisely the effect that I’m looking for, but am confident that I can still come up with something interesting. To date, I have only been able to produce the following – a disappointing monochrome mess.