Goals of this project:
Use the experience from designing a scanning key matrix for the DrumFighter prototype to create a Wicki-Hayden based chromatic instrument with more buttons than there are pins on the Arduino Due.
Combine both of the separate button decks from the Melodicade Prototype #1 into a single more portable device.
Add options for expression via a breath controller.
Throw everything-including-the-kitchen-sink into one device so loaded with inputs that the control keys don't even fit on the top deck.
What was learned from this prototype:
Throwing everything-including-the-kitchen-sink into one instrument makes for a difficult and frustrating playing experience.
Blowing into a differential pressure sensor to play notes is exhausting and not very fun.
For harmonizing, playing button decks side by side is far easier than layering left and right hands on top of each other.
Aerosol clear coat is way easier than epoxy resin for sealing in paint, but the awful smell persists for weeks after application. Maybe painting PLA isn't feasible for devices that need to be handled like this.
I had fun building this, but playing was a bit of a frustrating experience.
Frustrated with the lack of analog expression from digital arcade buttons, I tested a differential pressure sensor to change key velocity via breath control. This worked amazingly well, but even with a 3D printed mouth piece, blowing through a latex hose is awkward, and smells unpleasant. Never having seriously played woodwind instruments, I also disliked being light-headed constantly. Without changing to a different input medium, maybe foot pedals are the only satisfactory option for expression changes.
Having control buttons running the lengths of both sides meant that buttons were inadvertently being pressed all the time (octave changes, program changes, etc.), causing frequent controller resets to undo whatever random thing had just occurred. I liked the idea of an entire 6 octave instrument easily fitting into a backpack, but so much had to be sacrificed to fit 114 arcade buttons inside my 3D printer's 300mm² build volume, that it wasn't worth the trade-off. This prototype was destroyed to harvest the buttons less than two weeks after completion.