Antialiasing in Video Texture

This topic has 4 replies, 2 voices, and was last updated 6 months, 3 weeks ago by Thomas Fabini.

Viewing 5 posts - 1 through 5 (of 5 total)

Author

Posts
2025-04-28 at 9:36 am #81532

Thomas Fabini
Customer

Hi guys,

I am having some issues with a (dynamically loaded) video texture, where a client reported that the content of the video appears jagged when the viewing distance to the video increases.
The video is loaded to a texture placeholder png which had the Anisotropic Filtering set to 4x in the shader. In 4.8.0 this seemed to work ok, while in 4.9.0 I haven’t been able to see any effect when changing Anisotropic Filtering. The filtering texture in the Shader Editor is in both versions set to ‘Linear’.

Were there some significant changes to Antialiasing in 4.9.0 compared to 4.8.0?
Is there a setting I might have overlooked or something I missed?

Thanks,
Thomas

2025-04-28 at 1:02 pm #81536

Alexander Kovelenov
Staff

Hi Thomas,

The anisotropic filtering requires texture mipmaps, which in turn are quite heavy to generate each frame. That’s why both are disabled for video textures.

Strange thing is that you see degradation of quality from 4.8 to 4.9. Have you updated the engine in the app (check the browser console to verify)?

If you need to improve quality of your renderings try increasing pixel density with set screen scale (try 1.5).

Soft8Soft Tech Chief
X | FB | LinkedIn

2025-04-29 at 12:05 pm #81554

Thomas Fabini
Customer

Hi Alexander,

thank you very much for explaining this.
The two versions were indeed running 4.8 and 4.9, one wasn’t upgraded and they were showing the respective version number in the console.

I tried to further investigate the visual differences in between the versions and – as you assumed – came to the conclusion that there isn’t a degradation of quality in between them:
The placeholder texture was probably set to “repeat” in 4.8 as opposed to “clip” in 4.9.
This lead in 4.9 to jagged edges around the video frame – but the contents of the video look about the same (as far as I can tell).

Does Hidpi actually apply to VR projects as well?
If so – wouldn’t this have a considerable impact on performance and framerate on headsets?

2025-04-30 at 12:09 pm #81574

Alexander Kovelenov
Staff

No, HiDPI should be ignored for VR sessions, but rendering on the devices themselves is quite slow because of so many pixels on the screen. So one should optimize the apps carefully.

Soft8Soft Tech Chief
X | FB | LinkedIn

2025-05-05 at 3:03 pm #81759

Thomas Fabini
Customer

Alexander Kovelenov wrote:

HiDPI should be ignored for VR sessions

Makes sense, thanks for the feedback Alexander.
Author

Posts