I very much would like to spot test textures using small sections of a large model; I've created a few very small test meshes for that purpose and have imported them in RC.
The problem is that even for a 512x512 single tile on a tiny 1,500 triangle mesh subset, RC is taking ~5 hours to process on 32 cores, consuming ~50GB of RAM out of 128GB total. That's about the same amount of time as it takes RC to compute (12) 16K tiles for a 20M triangle mesh in the same project.
Yes, the project contains around 4000 images -- but only a dozen or so images relate to the test mesh, so something feels wrong here.
Does this behavior square with your experiences?
Is there any way to compute low res (~1K) textures for small meshes without incurring the same computational penalty as a very large model with a very large set of texture tiles?
Please sign in to leave a comment.