BeatBus is a experiment both in turning passive waiting areas, such as the buss stop, more alive with interactive music as well as a try at finding new interesting ways of mapping music within a small environment. What manners of limited musical composition can be made without previous knowledge? How can the interest of a quickly passing commuter be captured? These are some questions we ask ourselves while we practice experimentation in search for the new style of social music creation.
The software is begining to fly! We have now built a breadboard prototype and connected it to an arduino mega. The red LEDs are hardwired to the buttons while the blue LEDs are controlled from Arduino outputs. Here the arduino is running a simple testprogram that has the blue LEDs high by default and sets them to low when pressing the corresponding button:
And here we've managed to control the LEDs from Max/MSP! Button input into Max has also been confirmed to work (not shown here):
In other news, the plexi-glass discs has arrived! And today we decided on these lamps with this power supply. It can deliver up to 41 A when having a 12v output, which should suffice. And then some. The lamps were chosen so that we can be assured to avoid fire hazard as they are safely (?) housed.
Wednesday was much about testing lamps. Our biggest concern with the halogen lamps is the heat they emit, but since we did not know the temperature of the lamp, nor how much heat wood stands, we did a day-long experiment.
A 19mm hole was drilled in the wood, where the lamp was inserted and held back by taping the wires (soldered to the lamp) onto the piece of wood.
We once again made a temporary tile with some old plexiglas and a thin piece of paper to dim it. Then we measured the temperature. The highest temperature measured during the day was 146°C, using a 20W halogen lamp with 35 mm diameter, on 12 V.
During the day we also got a lot of help from Amir, who showed us Max/MSP that could be used for mapping Arduino to a synthesizer. For testing this, we started to make a miniature floor with buttons as tiles on a breadboard.
The bigger part of the Monday, we got into the practical work. We did one small prototype in cardboard to test how a connection could be made and to test the foam-as-springs idea. We also found a piece of plexiglass and tried to find the best way of lighting a tile up.
The biggest change from last week is that we realised that 32 tiles would be too expensive and a bit too much to make during the time of the project. We settled with doing 12 tiles instead, 3 times 4.
Cardboard prototype
Per took initiative to make a cardboard prototype, and was soon surrounded with cardboard, foam, tape and copper tape.
When the tile is pressed, the strips of copper tape get in contact, and the multimeter beeps
Illuminated plexiglass tile
We experimented with different lights: Small pocket lamp, RGB LED and 10W lamp. We also tried to cover and uncover the plexiglass with different material, to get the prettiest shine.
The idea with the tin foil is to spread the light...
... although it is easier to see the light when the glass is blurred.
The LED tend to make a spot, and not to spread...
... so we tried to block the spot with some tin foil...
... but then we did a little "spreader" in paper.
We realised that one single LED might not be enough for illuminating a whole tile of 30x30 cm, so we are now considering to use 10W lamps instead, using 12V power supply.
Today's conclusion At the moment, we are aiming to make 12 tiles. Pre-cut plexiglass (30x30 cm) has been ordered, at a cost of approximately SEK 1000, after we found out that wood and plywood already exist. Lamps are yet to be ordered.
So, off we go. Today was a lot of talking to and fro about the how the implementation should be designed. Really feel like we got going in a positive direction. Three main areas where analyzed:
electronics: How will all the wiring be made? Will we use shift registers? How many LEDs?
Non-electronics hardware/Tiles: How should the tiles be constructed to allow construction of as many as 32 tiles in the short time we have for the project? How can it be made robust. Will it look nice?
Conclusions
The idea of how this should be realized is currently this: We first make some experiments on how to best construct single tiles or 2x2 tile sections. Using conclusions from this we try to come up with a scalable solution for 4x4 tile sections (two at most) or maybe smaller if we think that will be too time consuming for the limited time we have.
In the meanwhile, we should try to make a final electronics solution and try to make some small test cases for testing the software. Also, we should come up with a design for the software implementation. Specifically, we should decide on the features the final prototype should have.
