design – code – interaction

For the presentation during friday in the second week we finally came up with name and logo for our project:

Presentation went ok. Our concept was quite strong but the only prototype we could show was this remote SLR control. Not very impressive for two weeks of work comparing to other groups…

Third week started and we didn’t feel well with using SLR Cameras for our prototype. They are expensive, quite big and heavy. We really wanted to created something even more tangible, something that you can play with. In face of lack of both time and skill to build our own camera we finally decided to go for webcams. Ujjval bought some old / broken webcams on the flee market during weekend and started to test them immediately.

We hadn’t enough cameras so we ordered 5 more – the cheapest we could find. They were delivered two days later and this was a good time to move to the electronic lab.

At this point we decided that we are going to create 5-6 cubes/boxes with one button and one flash light and put webcams inside. We couldn’t use camera buttons for triggering image capture because we were using Macs and macam drivers and processing wasn’t able to handle that. So we needed USB cord for video data from camera, and some additional wires from button and light to Arduino board to control the electronics. We even considered putting everything inside one fat cord but eventually we left USB cord alone and used tape to stick other wires to it.

Buttons test:

We couldn’t find any strong white LED’s so we decided to use red ambient LED that could be visible through white milky acrylic.

Next I focused on programming side. A here things started to complicate. First we found out that using more than one Camera of the same brand connected to the same computer is not as easy as you could imagine. Most drivers don’t support that. And even it they do you can’t capture from all of the cameras at the same time. What’s more Processing can’t use more than one camera of the same brand at all because it refer to them by name which is usually the same for all connected webcams. Slightly better situation is in OpenFrameworks where you can refer to the camera by it’s id. But still we didn’t manage to connect more that two of the same brand.

It was real driver hell. I spend like 3 or more days switching from OS X to Windows, from Processing to OpenFramework and restarting my computer every time Processing sketch crashed (what happened many times). Often coding until late in the night.

At the same time Ujjval started making boxes. First prototype was from cardboard.

The second one was from transparent acrylic. This time we used laser cutter for the first time. It’s awesome!

Ujjval is real master when it comes to hand work.

Additionally because of his architecture background drawing models schemas for laser cutter was super easy for him so mass production of the boxes went very quickly. We even designed our own micro shield for Arduino to easily connect all wires from the boxes to the board.

Although Ujjval did most of the hands work I also got my hand dirty. I tried to cut boxes from the cardboard and do some soldering to have a break from programming.

Final test before exhibition: