This post is a warning to all Lumenrt users. There is a bug in the Microstation Connect 14 update in the exporting of files to Lumenrt 14
Objects, solids or surfaces, that have been symmetrized do not render correctly, apparently owing to a problem with surface normals. I have logged this and have been informed by Bentley that the problem has been identified and solved, but will not be available until update 15 in about six months....
Pending this, the only workaround I have found is to uninstall 14 then make a clean installation of 13
PDF
I hope this is of help to Lumenrt users
Thanks Max, that was the first thing I tried however because of the size and complexity of my models they will not export from V8i but will export from Connect edition.
Impatiently awaiting the release of version 15 of both lumenrt and connect.
It's a fairly constant battle - getting something to work with CE> LRT 14, but then discovering that something else doesn't. In my case, the mirrored objects problem is solved using v8i>LRT 14, but then this problem reappears; https://communities.bentley.com/products/microstation/microstation_visualization/f/visualization-forum-1977638070/194241/lumenrt---solar-reflection-on-ground
Exporting from CONNECT 15 to LumenRT 14 and the problem persists (making CE 15 unusable for this purpose). Do we need to wait for LumenRT 15 for this problem to be fixed?
See below; these buildings are all made from the same geometry, but mirrored elements show up too dark. Applying materials makes no difference.
Max,
When you mirror meshes I think the normals also get reversed. Check the normals on the darker meshes. It may help to reverse the normals on those meshes.
Thanks Ron - in my case, these are Smart Solids. I've not found a way to flip their normals.
As Martin says, it would be impractical to search through a large model for flipped Mesh/ Surface normals (even if this worked). This renders fine in Luxology/ Vue and one would expect LumenRT to handle all geometry in the same way.
You might be able to convert the smart solids to meshes and then reverse their normals.
I know it is not ideal, but it may be a workaround until a better solution presents itself.