Pure awesomeness

 From:  Michael Gibson
5880.4 In reply to 5880.3 
Hi Claas,

> Would this in theory than also mean that in Modo you should not edit the
> mesh or work with it or would Modo now handle the vertex shading data the
> same way as MOI.

You have to be in "item mode" when you're transforming the mesh in Modo, you can't move individual vertices around or else you won't be able to keep using the vertex normals map.

Also occasionally Modo may have a problem with triangulating certain complex n-gon shapes, in some situations it can end up making a thin slivery reversed triangle along certain sides of certain n-gons. If you run into that issue then you can instead export that particular model as "Quads & Triangles" out from MoI and that will use MoI's n-gon triangulator instead which can do a better job since it is optimized to try and avoid making degenerate triangles when it converts an n-gon into triangles. It's pretty unusual to need to worry about this though.


> I did a test with STL meshes, and noticed that the same issue persists,
> however the RHINO model comes in fine only with few mini issues. Does Rhino
> prepare the STL model differently? I am just asking because I am genuinely interested
> in understanding what is happening behind the scenes.

Most likely the same artifacts are happening on the Rhino generated mesh, but you probably generated a more dense mesh from Rhino with smaller polygons in it. When you make smaller more evenly sized polygons it can help to reduce the shading artifacts from this problem, but it will not totally eliminate them. But basically the artifacts will be more localized to smaller areas of the object rather than stretching across large areas.

You should be able to replicate the same kind of result out from MoI by using the "Divide larger than" setting to make your output mesh diced up into smaller more even sized polygons rather than having some polygons in it that may stretch a long ways across the object. You put in a distance for "Divide larger than", and any polygons larger than that size get subdivided. So for example if you have an object 10 units across try putting in "Divide larger than" = 0.5 and that will dice the generated polygons up to be smaller than 0.5 units in size.

That's not really a complete solution though, there still will be problems that will be particularly noticeable if you are using reflective materials or doing close-up shots. Again I'd really recommend using a different renderer that can use vertex normals, it makes for a much higher quality smooth shading when the proper vertex normals get used. Blender is the only rendering program that I know of that is not able to use them.

- Michael

EDITED: 11 May 2013 by MICHAEL GIBSON