PHYSICAL COMPUTING

Once I had the prototype working, it was time to get serious. I secured an installation space at Squeaky Wheel (712 Main Street, Buffalo). The room that I would be using measured almost 10' 10.5" x 18' 6", which was fine in terms of width, but perhaps a little long, considering the width of the projections.

The door to the room was centered in the wall, so I decided that I would build a temporary wall in the space, placing the entry to the installation to the side. This wall would limit the length of the installation to 12'. The size of the room allowed me to determine the maximum number of buttons in the room, which would have been 120 (based on 1' squares placed side by side). Eventually, this number was reduced to 22 buttons; aside from 120 buttons seeming like overkill in such a small space, I also didn't have easy access to 120 ASCII characters via my hacked keyboard. As I started to construct the floor, another issue arose that also lead to reducing the number of buttons: if the panels were placed side by side, there would be no negative "safe" space in the room where people could stand without interrupting the flow of the video (standing on a panel effectively paused the video, constantly returning the movie clip to the second frame).

For aluminum panels, I used recycled plates from a printing press, cutting them into 14" squares to minimize waste. I drilled a small hole in the corner of each square, where I attached the wire that ran to the breadboard. I laid them out in five rows, alternating the number of panels (4,5,4,5,4). This seemed to be the optimum layout, assuring that panels would be pressed as people walked through the room while still allowing for negative/safe space (columns were separated by 12", rows by 9").

After laying out and wiring the bottom matrix, I covered it with carpet padding. I cut a holes in the padding centered over each panel so the the top and bottom pieces would connect when stepped on. Since aluminum is a soft metal and will hold a shape once bent, I also placed a small piece of padding in the center of each hole, aiding the prevention of creating permanent concave bends in the top panel. The photo below is a shot of the padding with five small holes cut for each panel (this didn't work so well).

With the padding in place, I laid out the top panel matrix and wired it.

Finally, I covered it all with a piece of carpet, creating a smooth walking surface. Considering several restraints (time, projector availability, finances, and Flash movie width), I temporarily discarded the fourth wall (though it will return in the final installation).

With the physical interface constructed, it was time to return to the Flash file. I created additional movieclips to total 22 and wrote 18 new lines of poetry that would (hopefully) make sense in any configuration. I increased the stage size, mentally dividing it into thirds: video grid wall, text wall, and fullscreen video wall. I expanded the original video grid to echo the new floor, laid out the new text movieclips (now limited to 10 seconds), and placed full-sized instances of the video clips. In the actionScript, I added lines in the keyListener to include the new symbols available on the floor, modified it to activate both video and text instances, and built an array using the keypresses that activated the video and text on the other two walls.

_global.vidArray = new Array ()
_global.lineArray = new Array ()
onEnterFrame = function ()
{
currentMovieClip = _root[vidArray[0]]
trace(typeof(currentMovieClip));
if (vidArray.length > 0)
{
if (currentMovieClip._currentframe == 1)
{
currentMovieClip.gotoAndPlay(2);
}
if (currentMovieClip._currentframe == currentMovieClip._totalframes)
{
vidArray.shift();
currentMovieClip = _root[vidArray[0]]
currentMovieClip.gotoAndPlay(2)
}
}
}
keylistener = new Object();
keylistener.onKeyDown = function() {
if (Key.getAscii() == 113) vid1.gotoAndPlay(2) and line1.gotoAndPlay(2)
and vidArray.push ("v1") and lineArray.push ("l1")
if (Key.getAscii() == 119) vid2.gotoAndPlay(2) and line2.gotoAndPlay(2)
and vidArray.push ("v2") and lineArray.push ("l2")
if (Key.getAscii() == 101) vid3.gotoAndPlay(2) and line3.gotoAndPlay(2)
and vidArray.push ("v3") and lineArray.push ("l3")
//these lines continue for all 22 panels -- i've cut them out in //order to shorten the page
if ((vidArray.length > 2) and ((vidArray[(vidArray.length - 1)]) == (vidArray[(vidArray.length - 2)]))) vidArray.splice((vidArray.length - 1),1);
if ((lineArray.length > 2) and ((lineArray[(lineArray.length - 1)]) == (lineArray[(lineArray.length - 2)]))) lineArray.splice((lineArray.length - 1),1);
}
Key.addListener(keylistener);
//---------------------------------------------------


I added an enterframe function to call instances of the movieclips on the third wall to play, then discard the first member of the array (the clip that was just played). This effectively built an automated editing system in which the user's path through the room created a new movie pieced together from the activated clips.

The sound in the space will respond to light levels in the room, from both the video and the shadows that people throw on the wall as they interrupt the projections. To measure the light, I used a photocell resistor wired into an Arduino board.

I found several helpful resources on the Arduino website, including the following code that will send pin information from the board to Max/MSP:


//---------------------------------------------------
#include
/* Analog/Digital inputs to MAX/MSP trigger
* ------------
* send serial values to MAX/MSP to trigger something
*/
char firstChar;
char secondChar;
void setup()
{
Serial.begin(115200);
}
void loop()
{
if (messageBuild()) { // Checks to see if the message is complete
firstChar = messageGetChar(); { // Gets the first word as a character
if (firstChar = 'r') { // Checking for the character 'r'
secondChar = messageGetChar(); // Gets the next word as a character
if (firstChar = 'd') // The next character has to be 'd' to continue
messageSendChar('d'); // Echo what is being read
for (int i=0;i