Tweet
Hey guys, I am trying to narrow down the factors that determines a files size. I know that the Texture size that is generated from the "Textured Mesh stage" determines the size of the .jpg and .png of the texture file. Most of the datasets are about 350 photos and their file size tends to be about 500mb. Since generating numerous point clouds and meshes on such numerous datasets can be quite time consuming can anyone give any advice that can reduce the file size.
-
-
As a first quick suggestion, you should use binary formats and jpg instead of png. For example for the mesh, fbx or binary ply are better then ascii obj or stl. For point clouds, I suggest you to export in e57 or las.
Then you should consider also to use decimation / retopology filters.
Here's a nice tutorial about decimation and density vs quality in general:
Comment