How do I use occlusion area correctly?

I have read about this but I can't figure out the concept of how to actually set it up properly.

My scene is 99% empty and will be filled by moving prefabs once it's running. Some prefabs make a tunnel, some just walls and floor. Now, how should I set it up so only object behind solid walls are occluded correctly while everything is moving around?
Which objects should have the component on them, only one or several or should it be on a static one?
What area should be covered by the component, the whole thing where the prefabs will be moving or just part of each prefab?

 

bump

 

Bump, is this feature that obscure?

 

bump

 

Ah I see, thanks for the info.

Just to make one clear, when you say "where the camera to be" you actually mean its position and not its far plane point and everything between?

 

Thank you.

 

What does a Occlusion Area do if "is view volume" is disabled?

 

I cant work out what an occlusion area does, can it be used so that if you look out of a window you dont see any of the geometry out there? It doesnt seem to do that for me. The only way I can achieve this is with an opaque renderer outside the window but then you cant see into the skybox which is what I want. Spent 2 hrs messing, very confusing.

edit: Occlusion portal works as expected, I got confused between them, still not sure what an Oclusion area is though

 

I'm missing information about light and Occlusion Culling.
And I assume that light behavior cannot be controlled in any way.
Light marked static or dynamic makes no difference. (I assume I cannot disable light from being an occludee)

What about occlusion culling and transparency, right now it looks like setting object to transparent will disable object from being an occluder. What if I want my transparent object to be occluder ?

 

When you don't add occlusion area It is created by default during baking process. It encapsulate all scene.
But then you have a big occlusion tree (and memory usage) for algorithm to traverse.
But when you know that your camera will never trespass some regions you can minimize this burden by placing your own occlusion areas in places where your camera can be.

 

When I place some Occlusion Areas what will happen when my camera is outside this areas. Since as documentation says volume around scene is created only when there is no Occlusion Areas in scene.

There is also a question. If I got a scene where camera can free-fly everywhere is there a point in creating Occlusion Areas in some places ?

 

For what I could find out, if you have an area on a part of your scene it will fill out the other parts but you would probably want to to put areas even with free-fly camera

"Unity performs higher precision calculations when the Camera’s position is within a View Volume"

That means, areas help make the occlusion better at the places you put areas in (for example in a free world like let's say a Minecraft server, you would want it in the lobby of your scene to make it preform better but for the other parts you don't need it as much).

Hope it helps, I am in the process of making a tutorial about those things so I will share it here when I'm finished

 

So what it is doing under the hood so it performs better than scene without Occlusion Area ?

 

lol still cant get these to work,

thanks I finally understood it today.

 

Hey,
I just made a tutorial that explains everything you need to know about how to use the occlusion culling
Let me know if this helps you
Link to the video

 

can we define some arbitrary convex volume for occlusion area, my level is consist mesh generated at runtime from point cloud data?