Yeah these things do pack a lot of useful technology into something so small and cheap..
I was recently watching a vid were some chap was gutting one to use it’s key parts in an RC drone. The accelerometers are good for the auto stabilisation hoopla.

Back to midi however.. Is a bespoke hardware mod/device really needed for Wii/Nunchuk midi control? Im thinking not.. Simply because it is already very cheap and easy to use a Wii controller, as a midi controller. I’ve been doing just that for a couple of years now.

The regular Wii controller has bluetooth, obviously.. And these days many computers have it included also (if not, then it is cheap and simple to add). So then all you need is software to translate the wii movements and buttons, into midi messages.

Macs have Osculator (which is great!) www.osculator.net
And PC users have GlovePie, which does a similar thing.

Im guessing you plan to do a computer free/standalone version though? But when it is already this cheap and easy to do (and we all have computers) im not sure how justified it would be to do it standalone..

@Luap: you are right with your point, but still… what about people who prefer to work without a computer? On many scenarios turning Wi-fi and bluetooth off can increase audio performance or solve drop-out issues so using these kind of devices is a bit of a hassle… just o say that there is plenty of reasons why a device like this could be great!

Osculator rocks.

What I plan to make will primarily interface with my next project (think of it as a “Nunchuk programmer”) ; but I’m wondering whether I go through the extra effort of adding 5 or 6 extra parts on the board and 100 lines of code to make it work as a standalone unit.

If you’re doing it anyway, then yes, add the 5-6 parts and code to make it standalone. That said, I’ve not been a huge fan of all this alternate controller stuff. Probably just stuck in my ways, but there you go.

I’ve seen them taped to guitars to control efx. Mostly it should be computer free,
Buttons trigger on sequences/arpegiators/portemento on
accelerators doing filter/vca
some sort of force to scale with octave ranges limited to 1/2 octaves of midi note ranges, going wider would simply be un trainable for live work so put some user designed sysex loaded limits that can usb off software one time, pre-use.

I have built one and coded 75% of the firmware. Buried for the moment. It has simple modes like playing MIDI drums or sending CCs, but I had planned other modes to control pads or arps. Too gimmicky and too unreliable for performance for my likings…

Re. the Kinect thing, you have to do a lot of processing on the depthmap image to get useful data from it. On a desktop computer, you can offload that kind of stuff to the graphics card (essentially a massively-parallel computer in it own right, with specialist DSP for dealing with millions of pixels), but I imagine dealing with the kind of data-throughput and image-processing needed would put it way beyond the capabilities of the kind of cheap multi-purpose programmable chips Olivier works with. I’m not putting them down, it’s just they’re not designed for image-processing duties.

a|x

Well the doppler effect approach would be cheap. You need a ultrasonic transmitter and a receiver.

@Pichenettes the Kinect computes a depthmap image, but doesn’t do anything else with it, you have to do the processing yourself.

a|x

PCB was extracted from WM+
During soldering TWI cable i’m assidentally knock off 0402-sized capacitor near IDG600 gyroscope chip (C12 at the bottom). No luck searching datasheet. No luck using .22uf like in IDG300. I’ve desoldered identical one, soldered legs to it. Insert into multimeter – and measured .1uf =) Than hours of searching for broken cellphone pcb, its schematic and needed capacitor, desoldering, soldering etc. Next time i’ll be more attentive. My son would never forgive me cannibalizing of second MotionPlus :D
Next i’ve enabled it’s internal 3.3V regulator with wire jumper (Arduino Nano not able to power 3.3v devices without usb connection, the regulator is inside FTDI chip)
Connected unused Shruthi MIDIout pins to gnd + 5v to power controller by MIDI cable (later realized that it was bad idea. or i just need decoupling caps)
Added midi activity led + two buttons, one for “Note on” another for gyroscope recalibration.
And here it is in action
Hardware is finished. There’s still lot off things to do with code (first shrink it to fit atmega8, i need my Arduino Nano back =) but i’m very tired.