I did some more experiments.
Minimum value is Fusion 0.13, minimum value mesh construct is 0.05
Then generally tolerable results are obtained.The “Edge Trimming” feature is really missing
If you fused at 0.15 and meshed at 0.05 then you induced artificial points, beyond the structure of you cloud point , this do not add anything to the final details or accuracy and can diminish the details, edges or as more noises, that are principles of 3D modeling.
You construct a cube from 8 points, adding 16 points or subdivide it will not make it any better.
If you fuse at 0.15 you need to mesh it at Grid of 0.15 to preserve the original point cloud at its best in accuracy and details as that will be your maximum.
I suggest you follow the rules of the workflow for best results .
1 Like
Yes, I totally agree. The task was to figure out the limits. For example, I found out that parameters less than 0.15 give a lot of noise. But at the same time, mesh 0.05 does not add noise with this optimization.
I guess you still don’t understood what I was talking about .
Meshing it at 0.05 give you nothing but additional artificial points , the main cage is still only 0.15 mm and so it’s resolution. You can’t make a scan better than the point cloud. But you can make it worse meshing it the wrong way as that can kill actually fine details like corners or edges.
Testing and learning is good practice, but taking sometimes advice from someone that know well this device will speed up your task mastering it to perfection in place of guessing.
I understand exactly what you’re talking about, thank you very much! But since these scans are not orders, I’m curious to see what happens if I try and experiment. It’s clear that you shouldn’t do this at work.
1 Like