Hi,
We have a VR scene that incorporates either an integrated mesh or a 3D object scene layer of an area imported runtime from an enterprise portal.
The model is derived from photogrammetry generated by a drone survey.
The problem I'm encountering is with the Level of Details (LOD) management. When using the integrated mesh, the more detailed tiles don't load at the appropriate time (see video attached).
We also tried using a 3D scene layer with a lighter model. In that case, there's no LOD management at all. The model appears clear (see screenshot attached), but this results in significant performance issues.
Is there a way to manually control the LOD? There is a specific way to publish the model on the portal to have it working properly in Unity? How does the LOD of 3d models function behind the scenes?
Many thanks!
Hi,
That does look much coarser than I would expect.
A few questions:
Thanks
Hi Michael,
thanks for the reply.
Following the answers to your questions:
In the VR sample it seems that the LOD visualization works fine with both 3d object scene (San Francisco) and integrate scene layers (Girona). Maybe is there a specific workflow we need to follow when we publish the model as slpk?
Many thanks,
Ferdinando
A couple more questions:
Thanks
Hi Michael,
please see my answers below:
The integrated mesh slpk, as the one I showed in the video, comes directly for the photogrammetry software . However we usually need to apply further refinement and optimization to the model using 3d modelling software such as Maya. In that case the workflow is: Import FBX in CityEngine > Manually geolocate > Export to File GDB (Multipatch) (gives better results than export directly to slpk) > In ArcGIS Pro from the MP create 3D Object Scene Layer (since Create Integrated Mesh Layer is only supported for OSGB format) > Publish
From fbx we can only obtain 3D object Scene Layer with this workflow. Is there another workflow to create integrated mesh from fbx?
Many thanks,
Ferdinando