The KinectController script has an option to use a KinectV2. I disabled the Azure and gave it a try. It’s plug-and-play baby! The Cubeman object knows to disable joints that aren’t tracked in the older sensor and is able to infer the positions of other missing ones like the clavicles.
Something I’ve noticed is that the Azure sensor infers the positions of limbs outside the frame much better than the V2 Kinect. This solves the notorious collapsed legs problem.
A screen capture from my Inherit Avatar project showing collapsed leg joints
I successfully subscribed to the OnUserAdded and OnUserRemoved events in one of my own scripts. Now I can begin creating logic and building a proper player prefab.
Did that thing I always do where I duplicate and rename all the scripts for my new scene. This time, I finally used the Replace All button in VS Code. Smarter at work.
I paired my left and right hands to the Cubeman object and viola, I’m moving particles! But, since the camera Z displacement is 0, I need to make my particles wayyy smaller. Easier to do that then to scale and move the player, me thinks.
I had to rewrite some of my mesh combiner logic to get the SDF to generate with my new player prefab. In my previous scenes, I wasn’t actually using my OnPlayerAdded and OnPlayerRemoved events to add players. I basically had to come up with all that logic today.
There seems to be an issue with multiple users in the scene. I developed a dummy user for testing so I won’t need to bug my mom anymore. She doesn’t have time for this.
Only one sphere mesh is being added to the mesh combiner
Welp, it was an indexing issue. GOTTEEM [DONE].
