After some initial testing with the BioWave headset, it became apparent that I should be using a bluetooth serial connection rather than connecting the arduino to the computer via the USB port. The BioWave instructions suggest that I should use the sensor exclusively with their adapter. However, I thought that the price of this adapter was too high to be viable for use within this project. I also thought that since the software is based in Max, I would likely be able to develop a similar system in Max and Arduino. As the system is bluetooth based (to avoid distortion of the signal resulting from the mains power supply as well as the potential risk of user electrocution) I decided to build a similar, battery powered bluetooth system based on the Arduino BT. I had based much of my initial headset development via a USB serial cable, developing on the Arduino2Max code and patch:
To develop the bluetooth system, it was merely a case of uploading the sketch to a bluetooth Arduino and aping the device with my computer. Data from the Biowave could then be manipulated in either Max MSP or Processing.
However, further working with the project I found that it was difficult to analyse the raw data from the sensor. Much of my work with data visualisation has been based in FFT analysis of an audio signal. Due to time constraints and the relative ease and convenience, I used the data from the sensor to modulate an audio signal in Max MSP. As I was interested in working with the raw data, I decided that the signal should be used to modulate the frequency of a bank of oscillators, similar to the wavetable synth used in the Binauaral/Monaural Beat Generator. The audio from this biosignal controlled synth could then be routed internally using soundflower, and visualised in Processing like any other live audio signal. I decided that the user should not be able to hear the signal, largely due to the irritating noise it makes which could be seen as off-putting. I felt that this would be out of place given the immersive and dissociating nature of the piece.
In constructing the headset, the sensor was mounted into the frame of a pair of modified toy night-vision goggles. The electronic components were harvested from the toy for future use, before removing sections of the frame to enlarge the headset for adult use. The frame was kept in place (save from some trimming) as well as the elastic fitting strap. After giving the goggle frames a coat of matt spray paint, the sensor was glued in place with the electrodes fitting the brow of the wearer. The board, battery pack and excess wires were placed in a zip-up wallet with a reasonably rigid frame (bought from Pound Land), which was then glued to the side of the headset with a hot glue gun. This allowed easy access tot he board to turn the battery pack on and off and to replace dead batteries.The resultant headpiece makes for quite a convenient biofeedback interface for use within the piece. The bluetooth connection allows the headset to be used without wirelessly, whilst the whole thing is sufficiently small to be used alongside headphones. The whole unit (consisting of the frame, sensor, board and battery pack) is pretty light and easy to use.