I wrote something so I can see my hands in vr using unity 3d. At first, I mounted it on the desk, rendered a panio, and it sort of sucked. So I mounted it on my head. But I want to fix the depth and reduce lag. Code is in the sharevr project.
So I use the Tesseltation of the GPU. So I send the hightmap their and then generate the geomotry on the fly. It works perfectly, and will be in the sharevr project when I push. It runs at 90 fps with no lag.
- Lag Compisation
- Render in world space (and have it linger)
- ‘Stamp’ onto a 360 rgbd photosphere for persitance and better streaming (this sounds valuable)
- Use problistic voxels, render on gpu. (can easily do 1gb)
- Fill holes (becuse the black thing that I can’t see is proably just a distance.
- can render corection like a shadow. Almost…
- Could just do Dilation for the gaps.
Stero stops mattering at a few meters.
The new sensor makes a clener map. Would also be nice to have two recivers or led to remove the shadows.
Could I just mount two gopros?