Forum Discussion
Supersampling in Unity integration?
As some of you have noticed, the UDK OVR integration has the ability to render the scene to a larger size than the actual Rift screen size (using the ScreenPercentage config option), and then use that larger render as the input into the lens distortion so that the final result that you see has less distortion artifacts (like long streaky lines of the same color near the edges).
Is there a way to do this in Unity's integration as well?
Is there a way to do this in Unity's integration as well?
4 Replies
Replies have been turned off for this discussion
- AFAIK they are waiting on features in an upcoming Unity version to do this, including MSAA support in render to texture.
- As soon as MSAA on RenderTextures is supported in a future version of Unity (which is underway) this feature will be added. This will bring the fidelity of the Unity output to be on par with UDK and Unreal 3.
Best,
Peter "petergiokaris" wrote:
As soon as MSAA on RenderTextures is supported in a future version of Unity (which is underway) this feature will be added. This will bring the fidelity of the Unity output to be on par with UDK and Unreal 3.
Best,
Peter
Thank you very much Peter and Siggi! :)- Supersampling is perfectly doable in the current version, in combination with image-based antialiasing. Even without adding code, you could just bind a new rendertexture of a larger size to the cameras. MSAA would perform better for similar image quality, but you won't be able to use that with deferred lighting anyway so its not a universal solution.
But it looks like the stereo configuration is not exposed in the OVR plugin yet, so to increase the FOV to the edge of your vision as described in the main sdk docs you would need to do the distortion and FOV math yourself.
