Aw thanks so much Renata! 🩷

❤❤❤

aww thanks a lot! :) There are several different sounds that become activated or deactivated depending on the activity/inactivity of body joints over a period of time. Once a sound is activated, the position of the body modulates the individual parameters of the sound (e.g. right hand increases the oscillation frequency as it travels from right to left). Things like pitch envelope, delay time, filter resonance, grain size, panning, and frequency content are modulated via 20+ different interaction rules.

@@melochua wow that's insane! how does it detect motion?

@@xuanfilms :D I'm actually just using a normal web camera and then running it through a Posenet model to detect the joints of the body. From there, I get position data and a "score" for how accurate it is perceiving the skeleton (this is very useful for reducing unintended triggers). Then in Max, I made a small patch to measure the velocity of the joints, and in combination with the position data, I created the interaction rules for the sounds/visuals.

@@melochua well that sounds suuuper simple :P

Stephen Taylor Thanks a lot! Feel free to message/email me and I can explain it in greater detail if you want :)

shinjenny thanks a lot!! ☺️