Any missed steps here? How much do some of you simplify your captures?

I’ve been using a black out method with a turntable to capture my images and the results are great.

I wanted to know some opinions on how much to decimate objects for a 2.5D game. I know I can use certain visualizers in Unreal for textures but I also wanted to know opinions on vertices density. I have some experience with 3d but mostly I’ve mainly just been using reality capture because I sculpt things by hand.

So far my process is:
-shoot 30 images of object on turntable at a straight on angle
-then 30 more at a higher angle
-I then bring the images into Lightroom where I increase texture and clarity and adjust levels
-Then I drop those images into realitycapture where they I get good results.
-Reality capture
-Align images
-Scene 3d tools set ground plane
-Click set reconstruction and adjust
-Mesh model normal detail
-Texture
-1D Tools Simplify
-1D Reproject (click yes)
-Export top left: FBX Make sure normals are intact

Does this seem right? Am I skipping vital steps? It looks good in Unreal but I just want to make sure the files are as good and as small as possible.

There isn’t any specific secret workflow that might be looking for, you’ve pretty much got it spot on already from the sounds of things!

In future you might consider using other software for retopology (depending on budget, Blender is a good place to start). With a bit of work you can really condense meshes down to just a few triangles without losing too much detail, then reimport back into RC and reproject the color + normal map (this will give you more control you might be looking for)

Good luck - look forward to seeing any examples of your work :slight_smile:

1 Like