Hello community,
I’ve juste watched Valves video about advanced VR rendering and optimisation from GDC and was asking myself if there is something similar in Unreal Engine. Look at the video here : http://www.gdcvault.com/play/1023522/Advanced-VR-Rendering
I know there is the stereo rendering feature, but they mention other methods to gain framerate in the video :
- “Hidden Area Mesh” to early cull non visible pixels on the screen (maybe this can be achieved by placing a mesh just in front of the camera and enable early z-pass ? )
- “Radial Density Masking” to render the peripheral vision in a “virtual” lower resolution (rendering only half of the pixels in a kind of dither pattern, then reconstructing the missing ones with a simple filter)
- Assigning different “screen” zones to each GPU when having a multi-gpu pc
So my questions are : Are any of the previous features already in the engine ? How to enable ? If no already here, how hard would it be to implement them ? Is the any plan from epic to get this in the engine release ?
One other thing the mention is adaptative quality in real time (by analyzing the second last frame duration) to maintain constant framerate, but this can probably be achieved by blueprint or some coding.
Cheers 