If I understood correctly, my ratings must be of tolerance 12.00011 for example? I just tested a rating of 2.00 to 2.000001. Nothing changes in my STL control panel.
So I just tested what I had seen. The tolerance bar adjusts based on the overall volume of the room. Ex a 10x10x10 cube will be on the right in deflection at 0.00085915 mm against a cube 1000x1000x1000 = 0.08591522 mm
After I admit that it's still obscure for me this subject. I have a doubt about it, but if I understood correctly, larger pieces will have a coarser deviation in this context?
tell me if I'm wrong, but I think SW limits the tolerance for exporting files with a large volume and a high resolution. I don't know if there is a way to jailbreak SW, I looked briefly in the system options and document properties friends I found nothing. I thought about playing with the scales, but it doesn't change the resolution of the facets. Unless the slicer integrates it into its scale management. I'm going to look with the Meshlab software or something like that if I can refine the facets.
What I find surprising is having to go below 0.01mm not to see any facets.
The accuracy tolerances of 3D printers are, if I'm not mistaken, well above this dimension.
For me you should do a test with a large circle (depending on the size of the plate height 4-5mm and you make an export to the scale of 100 that you reduce in your slicer and the model with the right scale and then you compare.
What is sure is that on my ender 3 the facets are very little visible, or even invisible without having to touch this deviation parameter.
You also have to see the type of printer and if you have TL smoothers in order to smooth out the movements. (maybe the defect comes mainly from the printer)