Im Not sure if this is a BUG or Im doing something in the wrong way, help please

[ozukaru]

I'm mirroring uv by first making upper retopo mesh then assigning uv's then mirroring top to down geometry, but the final result show some imperfections that show that the mirroring process is not much precise. Could that be related to the fact 3d-coat is not mirroring the exact underlying triangle topology? note that in the upper the two triangles that conform the quad are not equally positioned as the bottom ones (btw the stretched checker patter is intentionally set that way in order to see the underlying triangle better). In the second image Im painting depth as an example of the problem, compare upper to lower.

Is that correct? is the problem related to the underlying triangle topology?  if this is a BUG will be fixed or is not that relevant? 

 

I tried mirroring using "general case", "Local Space", etc  all give the same result

 

PLEASE help, Im relying so much in mirroring uv in order to save resources because my models are for a game! THANKS.

 

 

Edit: added a third image to show what (I think) could solve the problem, an exact replica of the geometry (at a triangle level)

Edited October 10, 2015 by ozukaru

[Tony Nemo]

The top half has a seriously stretched UV island and no visible edge. Please show the unwrapped version for better trouble shooting.

[ozukaru]

The top half has a seriously stretched UV island and no visible edge. 

as I said that was intentional in order to show the underlying triangle and how the top it differs from the bottom one, I show it explicitly with the red strip. 

 

The real unrwapped is a typical overlapped island that means only one island is visible because both has the vertex in the same positions, but that does not mean that the underlying triangles are equally created. Anyway here you have it, as you can see seems to be ok. Thanks.

Edited October 10, 2015 by ozukaru

[Tony Nemo]

Sorry for my hasty mistake. I seem not to understand the problem as the last image seems without issues. What happened to the "triangles? 

[ozukaru]

Sorry for my hasty mistake. I seem not to understand the problem as the last image seems without issues. What happened to the "triangles? 

in summary Im having troubles with my mirrored uv's and im guessing is because 3d-coat does not seems to mirror the exact triangles topology (at least not in this case)

[ozukaru]

doing further tests mirrored triangle topology differences does not seems to affect uv precision if x mirrored but still does on y mirrored

My final solution was explicitly create the upper triangles before apply mirror on Y axis. In attached images, fixing the problem by doing this shows that the underlying topology DO affect uv precision so a correction to the mirror geometry tool is needed.

Edited October 11, 2015 by ozukaru

[Tony Nemo]

By "triangle topology"m are you referring to the Sculpt room surface? If so, it should play no roll in the Retopo room. There the polygons are independent of what triangles there were in Sculpt or should be. 

[ozukaru]

By "triangle topology"m are you referring to the Sculpt room surface?

It seems my english is worse than I thought

No, im talking about retopo mesh, it's clear in the attached images, the images with shaded model are from paint room though. 

 

btw I wish Andrew was by the forums like he used to do...

Edited October 11, 2015 by ozukaru