Our VR mode is not great though, needs work.
If you have a beefy headset you might be interested in trying this scene, which just blew me away when I tried:
superspl.at/view?id=9fd6...
Our VR mode is not great though, needs work.
If you have a beefy headset you might be interested in trying this scene, which just blew me away when I tried:
superspl.at/view?id=9fd6...
This is unfortunately quite common in scenes with shiny objects, especially cars.
I really hope "the researchers" find a way to improve this particular training artefact.
This is unfortunately quite common in scenes with shiny objects, especially cars.
I really hope "the researchers" find a way to improve this particular training artefact.
- no need to sort gaussians (which we currently do on CPU, terrible for large scenes)
- gpu now generates a z-buffer, so early out is possible and other z-culling approaches
Who knows, perhaps we could combine stochastic alpha and blended alpha...
- no need to sort gaussians (which we currently do on CPU, terrible for large scenes)
- gpu now generates a z-buffer, so early out is possible and other z-culling approaches
Who knows, perhaps we could combine stochastic alpha and blended alpha...