:||digital kitchen||:

Here's a description of our workflow and development of the project (sort of..)

Week 1:

After coming together to meet for the first time, Ash introduced an idea to create a musical instrument that would help people that aren't able to whistle making sounds. This involved blowing on sensors in order to produce sound.
Though firstly we approached the problem thinking and searching for sensor technology and possibilities, we tried to spare a lot of attention on the action itself: blowing and the resulting feedback, not only sound but also possibilities of visual feedback.
We discussed using paper attached to the sensors to trigger the bending as well as using flex sensors to produce the sound thru a piezo speaker and processing. We discussed having a visual element such as colored LCD screens that corresponded to the frequency of the tones which would be created in processing.

Finnaly we discussed about having pieces of paper attached to flex sensors: this would give the user a straightforward visual feedback of their actions and could (in a natural manner) capture the user's action (e.g.: blowing air pushes the paper forward, not blowing the paper moves it back to its original position: this movement is a faithfull transduction of the action.)

Week 2:

After agreeing upon our initial concept we set out to explore our system hardware/software needs:
3 data sources (sensors) + processor + sound engine :
3 flex sensors attached to the analog inputs of an Arduino board + serial connection to the ESS library on Processing.

Experimenting with ESS Library for Processing to get some sound out of a single flex sensor:
After struggling to get anything by generating a sin wave with the ESS library, we finally were able to get some sound, albeit a crudely mechanical tone. We were ecstatic to hear sound from our code, but still rather unhappy with what we felt was possible with ESS and our code. ESS gaves us a bit of troubles such as an unresponsive and delayed feedback from the flex sensors values and the generated sounds were far from our needs (too electronic and mechanicall changes in the frequency changes).
After presenting our results in class, the use of Max-Msp in place of ESS was brought up by another group and by Tom. We quickly latched onto the idea of having more audio control over the sounds we were previously getting. After a short demo in class of using Max we decided for next week to create a Max-msp patch that would generate sound from flex sensors data. Additionally, we purchased and mounted three flex sensors to a perf board base for support which would be ready for later use.

Week 3:

We presented our single sensor now producing sound from a MaxPatch we had created based off of another person's patch for a light sensor that produced sound. We adapted the patch for our purposes while hitting several road blocks along the way...
The three remaining things that we still had to address were:

1) getting Max to read from three sensors at once and scale their values

2) producing a sound that was more whistle-like, soft, gentle and less-electronic

3) assemble the instruments base;

For the sensor readings we had our troubles around max-msp but we finally solved a problem that was simpler than expected.(directly adding outputs to the receiver). To get more expressive data from the sensors we had to do some sensor values scaling which took a good part of our time.
We got some good tips and tricks with our patch from Eric B. and Jeremy and were able to produce sounds pretty close to our ideal, and though the distinction and scale of the sensor's tones were very close to one another their expressiveness has differences. As this was our first experience with max-msp and also with sound synthesis we had many difficulties trying to embrace many concepts at once but we also had very amusing moments doing sound synthesis explorations.
To built the base we decided to continue using paper as this material would support our desire of lightness;
An arc shaped base which would be housing for the circuitry but also the base object for the interaction instruments. This shape seemed the most appropriate as it can easily follow the users head movement.

We tested the product on the floor and got an overwhelmingly positive response. We were suprised to see how much fun and experimentation people used towards the instrument.
some of the user testing experiences gave us vital insights to important changes we still had to do in our project:
- most of the users were excited and quickly grasped our concep and interface approach;
- some users used the interface with hands and touch instead of blowing (may be caused by the papers shape?!) ;
- some users identified the paper's shape as tomb stones (negative!)
- more users requests: higher number of instruments for interaction = more papers and flex sensors with a musical scale; drawings within the papers;

user testing made us quickly change:: paper shapes to more rectangular shapes with rounded corners; covering the whole box sides..

For the future, we think that a finer differentiation in tone and note would be valuable so that the instrument had a consistency and reliability in note production that most instruments have. Also, as many users did, we do feel that a visual element would be valuable for our project.