Jump to content

AUTOLOD improvement


ua4192
 Share

Recommended Posts

Hi.

Would it be possible to add a special case for AutoLOD method when quality will be set to zero?

As addLODLevel is not working properly for imported meshes (complex meshes extracted from GLTF files) and instances.

For this reason, I am thinking to return to autoLOD method. This was working properly.

In fact my goal is by now just hide very small meshes the problem is that if you specify zero quality in autoLOD method the decimation process takes a lot.

Maybe would be a good idea for this case return a null mesh, this would be immediate.

COuld anyone explain me how to modify the code to add this special case?

Maybe the async decimation process should have an extra distance to be specified to return null mesh. This would be very useful to hide small meshes or very far away. Of course the first priority should be to  return add this LOD null mesh and after the others.

In fact in addLODLevel method you have this option, but when I try to use it it doesn't work properly.

Does it make sense for you?

 

Best regards

Link to comment
Share on other sites

Hi.

The problem is that I cannot reproduce the problem in PG.

My code is looking for the models in a DB using an AJAX call.

I cannot either add the models. Too many they and are confidential.

About using clones, I think tthat I should use instances. I want to take advantage of HW rendering and as far as I know, maybe I am wrong it is faster using instances than clones.

Anyway, I tried to use LOD levels with clones but it didn't work.

The steps I follw are the follwing.

1.- I make a query to an oracle DB and I extract the product structure. ASSEMBLY_LEVEL1 --- ASSEMBY_LEVELN -- PARTS (METADATA)

2.- I compute the different parts that will be used in my scene and I store them in an array of meshes with name key

3.- I start to import these different meshes but I disable them to be able to create instances later. I wait for all different meshes to be loaded.

4.- I build the scene instanciating the meshes imported in step 3

I cann add my code. but PG is not possible for me.

Best regards

 

Link to comment
Share on other sites

  • 2 weeks later...

Many thanks.

Finally I am using startdard addLOD feature.

As it is not possible to use LOD levels in GLTF importmesh function, I convert the same model with different decimation precission. I am using 5 different LODs (null,0.25,0.50,0.75,0.90) 

In this way it is working very fine, despites the fact that the time to load the assets (different gltf models in my scene) has doubled, after I use instancing and now it works like a charm

My final idea is to use a dynamic LOD distance based on the boundingbox size.

I am now able to see the scene I was trying to see ( 5600 assets with more than 100000 meshes instanced)

: )

I am now trying to make a kind of occlussion clussing based on the CAD logic of the plane I want to render.

I don't really understand why is so dificult to really see if the mesh is visible by the camera or not (hidden by other mesh). I

Many thanks for your support.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...