Smoothing Groups/ Normals ???

Playdo

I've got errors on a cog that I've made. I've cleared and re-applied smoothing groups, I've flipped/unified normals, I've exported/imported as an obj, I've remodeled, I've cleanly uvw mapped it, but still it's there.

Anyone know what it could be? It looks like a smoothing issue to me but I can't seemt o fix it. I'm using an Arch and Design brushed metal material, Max '10. I have another cog in the sene which doesn't have the same problem. I've uploaded the file if anyone's willing to take a look.

http://www.speedyshare.com/files/24917633/Cogs_Test_01.max

Defect:


Correct:

Ott

Please paint a giant arrow at what it is you think is "wrong".

r_fletch_r

have you tried doing a low threshold weld on your verts. it looks like there are overlapping verts in your mesh

Will Faucher

Ott wrote: »

Please paint a giant arrow at what it is you think is "wrong".

Are you serious? Look at the first picture and tell me how it is different from the second correct pic.

Honestly, I'm not sure what to say man. Why don't you just duplicate the correct cog and replace the faulty one?

Playdo

@ Ott: It's pretty obvious, it's all around the edges.

@ Fletchr: I have. Then I made a second model using Sub-d so there was no need to weld, but it still produced the same results. I thought it was a smoothing group issue, but then if it's sub-d modeled and meshsmooth is not linked to smoothing groups, it's irrelevent, no?

Did you look at the file? I didn't find any.

@ Prophecies: The correct cog's a different size to the faulty one, it needs a different number of teeth to fit.

r_fletch_r

didnt see the file. (was too occupied with the pictures ) will take a look now.

having looked at the file i think its an issue with the anisotropic reflections. have you tried a non anisotropic material?

Playdo

Ok, I sorted out the mottling on the original mesh through smoothing groups but there's still a distinct edge visible. I altered the anisotropy settings and it got rid of it but it alters the spec highlight. I couldn't do the same with the subd model though.

The weird thing is, when leaving the anisotropy settings as they were, I can get rid of the distinct edge by altering the smoothing groups but it just keeps moving across the model. It's making me question my knowledge of smoothing groups, but I know it's right ..when two adjacent polygons share a group, they smooth. If I put everything in the same smoothing group, I get the distinct edges everywhere.

Increased anisotropy:


Standard anisotropy - conflicting smoothing groups on corners (distinct edges as expected):


Standard anisotropy - same smoothing groups on corners (smooths as expected) - same+different groups where the distinct edge is: