I created a data glove to interact with a live audio granular synthesiser. I wanted to learn more about signal processing and physical computing. I was also interested to work with conductive material to capture the finger flexing.
- Circuit Playground
- Conductive Material
- ofxRapidLib for Machine Learning (KNN and Regression via Neural Network)
- Main granularsynthesis code came from Joshua Batty's Live audio Granular Synth
- Updated for OFX 0.9.4
- Updated Gui to ofxDatGui
- Install all addons in addon.make
- Use openFrameworks 0.9.4 and project generator
You can calibrate glove yourself or use some rough estimated values using the Glove Gui.
All button record test examples, train and run machine learning models automatically when toggled between active and inactive states.
- Static Hand - Record the hand flat so to help stop other gesture triggering
- Record Voice - Activates live voice recording
- Set Playhead - Control the PlayHead with x-axis on the accelerometer
- Manipulate Synth - Described Below accelerometer X, Y and Z are only used
To do this first set the sliders of the synth to your desired sound, you might want to record a sound first by pressing the live record button. After the your desired sound is set hold you hand in the desired posture and follow the steps below. Repeat process as neccesary.
- Regression Record - Will record how your hand is being held useing the accelorometers