Hi,
first thanks for the great software. I enjoy working with the software and already have some great results. However, I have a question regarding flat lighting. I´m working with a 90 camera rig and am doing Full Body Scans with success. The result is looking perfect in RC. When I upload to Sketchfab I have to choose unlit and vertex color off.
But what if I change the lighting? Like explained here https://support.capturingreality.com/hc/en-us/community/posts/360014944591-Suggestions-for-displaying-models-in-browser- The usual photogrammetry process would be to smooth the model in zbrush or blender or are there any other ways? Because isn´t it that smooting object will cause a loss of detail? And even if it doesn´t are there maybe other ways to do it. Maybe directly in RC itself? And what´s the exact problem that only flat lighting works out of the box?
Would be great if you could help me out.
Thank you very much
Hi drehmomenteDE,
thank you for your words.
What happen, when you don’t choose unlit? Flat lighting is the best for photogrammetry. Is there a problem, that model on Sketchfab looks worse? If you are uploading right from RC, the model has to have most 300k triangles, so this is quite big simplification, which means, that some of the details are missing.
You can smooth the model also in RealityCapture, there are tools for this. Of course, with this the details are missing, but you are able to loose some wrong model parts.
What do you mean by that only flat lighting works out of box?
hello,
are you looking for a PBR-Representation, to render scanned models in custom lighting situations?
The output of RC (RawScan) is only the starting point of this process.
After cleaning (and retopo) the geo, you have to generate some texturemaps to mimic materials, like albedo, roughness, normal, etc. to obtain a proper PBR-Material which reacts to lighting.
If you are interested in a more in depth explanation of this workflow, i am offering a consulting service…
best regards
Hi,
thanks for your quick replys. I think that PBR-Representation is the best term for what I mean. However, I do know the workflow (more or less) and did a texture reprojection in Blender with albedo and baking a displacement map. However I try to find a faster way of doing it. As reality capture has the ability to make a texture reprojection on a simplified model and generating a normal map I´m wondering whether a retopo is still necessary. And what about the uv-map in RC of the simplified model. Is this also not sufficient for a PBR workflow which lets the model react to different lighting situations? Hope I explained it well 
114781571112 - Thanks for the offer. It sounds very interesting. Despite the question above we will see if that´s an option for us.