Tuesday, April 22, 2014

Cheap materials and small surprises

One of the most rewarding and yet frustrating things about these projects is that you sometimes need to commit yourself to an idea before you fully understand all of your options.

For instance, my array of infrared LEDs and receivers (pictured below in all of its hot-glue-and-gaffer-tape glory), though effective, does not implement any of the various light tuning/focusing techniques I've discovered in only the last few days, far too late to incorporate into my design.

On the other hand, despite hours of research into rear projection materials and advice from fellow artists/technicians, the $3 white vinyl shower curtain that I'm using on the back of the glass panel as a projection surface works like gangbusters.

Also, despite my disappointment with the amount of tape, glue, and foam core that went into this design, the fact that the glass "wall" looks somewhat presentable (aside from the wood blocks clamped to the bottom, which will be painted) is a testament to the fact that sometimes it's only the surface (the illusion) that matters. If you can convince the viewer that you've made something seamless and solid, then it doesn't necessarily matter how much of a mess you're hiding behind that foam core!

Lastly, my plan for the projected visual material is to use macro-ish and/or textural photos from the space in which the walls are being installed (DL1). Here are a few of my favorites so far (I'm hoping to go even more abstract for the final versions):

Friday, April 4, 2014


Umm... Woah:


I guess putting micro controllers and sensors in balloons isn't so crazy after all.

Thursday, April 3, 2014

panGenerator - constellaction

This is the installation work I cited in my presentation today. In general, the group has a set of very interesting interactive installation other than this one. You can check out their website (pangenerator.com)


Wednesday, April 2, 2014

Lights on the walls

Just wanted to share some of my discoveries/research from the last couple of days.

First of all, the good news: I got some super cheap glass panels from the ReUse center. $5 each.
I went with 2 different kinds to see which I liked better. I'm not totally sure yet whether I like the idea of having 3 different walls or 3 identical walls. Hopefully if I chose the latter, they'll have enough of the same glass to chose from. There were definitely more of the smaller version sitting in the warehouse.

Also, research on AV forums has lead me to discover that cheap shower curtains actually make pretty decent rear projection surfaces. Not as good as the real thing, but for $3-6 a pop, I'll take it.

Now, the actual sensing is getting kind of hairy. Despite my delusional wishing, a standard infrared proximity sensor has an extremely narrow range (about an inch wide). I kind of expected that, but figured I should test it anyway. While the prospect of lining the side and top of the glass with dozens of these sensors is interesting, it's also too damn rich for my blood at $15 a pop.

So, my options as of now are the following:

Option #1: IR Receiver with an array of infrared LEDs

This option seems the most promising as these sensors have a wide viewing angle and are tuned to look for modulated IR light @ 38khz. It takes the job of detection off of the computer and only sends digital on/off. Plus, this is a cheap option! A pack of 25 infrared LEDs was $8, and each of these sensors is $2. I'm hoping I can get away with using, at most, 3 sensors per wall, pointing down.  Also, skin and hair easily reflect IR light, which I confirmed by pointing a remote control away from a DVD player and getting the signal back to the sensor by putting my hand and head in the light's path. The main challenge here will be modulating the LED pulses from an Arduino and getting enough current to them. There's also the issue of making sure that nothing else in the path of the light (i.e. on the floor) will reflect light back to the receiver. Glass is apparently very good at absorbing infrared (see Mythbusters).

Option #2: Infrared emitter/detector pair

I suspect that this is nothing more than the LED and receiver components from within a proximity sensor without the other circuitry. My concern is that this will not have the range or angle of view that I need, and I was not able to find this info online, but they were cheap so I ordered a pair to try it out. The benefit here is again doing all the detection on the Arduino with the added bonus of not needing to worry about modulating the LED.

Option #3: PS3eye

Back to camera detection. This is my least favorite option (most expensive, most difficult to setup in software, takes CPU resources, puts me back in Jitter land, might not be able to get 3 of these cameras working on a mac, etc). But, if I get it working, it's also the most robust. I could try to do things with location if I want. There's a lot of info out there about how to remove the IR filter from this camera, it's only $10 used at Gamestop (though I'll need 3 if this works), and I was happy to discover that the magnetic film from within a floppy disc (of which I was able to locate 2 in Groundworks) acts as a pretty good visible light filter.

Should have more to share before next Tuesday.

Also, on a side note, Chris Sies (the graduate percussionist) had a recital last weekend, and his parents were in town. Turns out his dad used to own a metal shop, and he was super intrigued by the Metal Mirror. He seriously sat and played with it for at least half an hour. It was great.