This feature is amazing, tested it on a few games and it works very smoothly!
When can we expect it to be released so all of our players can experience this performance benefit?
Thanks!
2 Likes
Yet we still can’t even seperate render distance from the graphics slider or use mouse side buttons lol.
Hard second, I notice no game-breaking bugs with this and believe it should be released to the client.
Will this work with Viewport Frames? Excited to see how it could optimize those, we could make actual hand held phone cameras, or a really good example is the tablet NVCS-3000 from DOORS, which had Viewport Frame lag.
2 Likes
Some experiences/ places are not working with Occlusion Culling as intended for unknown reasons. Engineers last I known are currently looking into those cases so I wouldn’t say its ‘good enough’ to move on to the next step for these impacted developers.
1 Like
Not doubting you whatsoever, but may you list an example of this? If it’s just stuff not culling, as in the example listed here,
It’s still better with occlusion culling than not.
[Studio Beta] Introducing Occlusion Culling - #51 by GhostyPep -Not really a part of what I am saying, but highlights there are issues.
1 Like
The first one is solved. The second one doesn’t really cause any major issues, It just means that occlusion culling will not work for that place. (I guess that also explains why it seems to work in Jujutsu Shenanigans but not Type://Soul). The truss thing is pretty minor and being fixed anyways, by the time they release it that will probably be fixed.
I get your point, but to me it seems like occlusion culling is not working at all in the second example, which honestly doesn’t change anything. The experience will just be as it is currently, while experiences it does support get the benefits of Occlusion Culling immediately for all players while the bugs are ironed out to support a wider variety of experiences. I might be biased since it actually boosted my performance very noticeably but I hope you at least somewhat see what I mean even if you don’t agree with it.
It changes everything when the OP is striving for a feature to work for everyone.
When Occlusion Culling is not culling… Then we have a problem, right? We cannot fall into a false consensus effect where it may work for you and yes, a vast majority, but there are those like me and @minhbelly where we were trying to wrap our head around why we are the special case in all of this and why our 17 million + tris scene weren’t receiving these amazing benefits.
To downplay it as an “oh, its just not working for them but it is for me and a hundred other people. 
oh, well? Doesn’t change anything.” actually changes a lot. It is an example of disregarding issues some developers may have, like me and perhaps the other people who liked the second post that might concur with their own issues with it; Something I am confident the ROBLOX engineers aren’t disregarding lightly.
3 Likes
Will there be the ability to create special “occluder parts” in the future that are invisible so they don’t cost any rendering performance themselves but still occlude objects behind them? It could be useful to disable occlusion on mesh based maps (since apparently complex meshes have occlusion disabled on them anyway, at least right now) and just shove invisible occluder volumes in the walls in between map chunks or in front of doors or what not instead, that way the occlusion itself wont take too much performance since you’ll only be checking against a small number of basic occluders, but should still get good benefits
6 Likes
I am not sure if this has been stated anywhere but will developers be able to exclude mesh objects, parts etc… from occlusion culling?
2 Likes
Could this apply to the faces of a mesh? For example, in a complex game with high-detail models, each mesh has many faces. However, not all faces will be visible at once because some may be blocked by others or by other objects in the scene. Even though the entire mesh is likely visible, only some of its faces can be seen from the player’s perspective.
1 Like
I was certain this was already a standard feature in studio, seeing this notification was kind of shocking. Feels like this is very, very long overdue and I really hope it gets out of beta soon because it’s so important to performance.
1 Like
It’s hard to explain but in certain positions it’s not culled and in most it is culled, it’s similar to how z fighting works and how certain angles may allow one face on top of the other. This is only an issue when far away from the object.
3 Likes
this is what I call the cube of invisible
1 Like
Why would you want to do this anyway? Just curious, since it only culls things that you cannot see in the first place.
1 Like
Does anyone know if Roblox has given an approximate release date for this wonderful function?
1 Like
Would it be advisable to not use StreamingEnabled at all and just use Occlusion Culling for a large dense map?
What is meant by that avatars arent affected? Aslong as its a model owned by a player or a model with an Humanoid?
1 Like
Occlusion Culling and Streaming focus on two different areas. Occlusion Culling improves rendering performance, but Streaming reduces memory usage and join times. Occlusion Culling does not “unload” parts of your world like Streaming does, it just stops them from being rendered, as though you had set .Transparency to 1. You shouldn’t use one as a replacement for the other, because they do different things.
1 Like