[Bf-committers] unwrapping on subsurfed meshes

[Jean-Luc Peurière]

Le 3 oct. 05 à 20:44, Nick Winters a écrit :

> this is not a problem with lscm
>
Where did i say it was a problem with lscm ?

> uv coordinates are interpolated the same way using and not using
> subsurf and simple subdivision

Yes and that is the cause of the problem. In a subsurfed surface,
the vertex of the control cage move to a position closer to the surface.
lscm (and others methods) is fed with the original positions of the
surface ( i dont speak and dont care at this point of the ones created
by the subdiv process). as the conformal map produced by lcm is
essentially an energy minimizing process, the result would have
been a little different with the final positions of the control vertex.

> however, subdivision surfaces will warp faces in strange ways.
> particularly if verticies don't have edges coming out at regular
> angle
>

The control cage is the main cause of the problem, as the  
distorstions of the
inner quads are by definition smaller. but it is true that they will  
cause
problem if the mesh is too irregular


> I think having a proper mesh is better

> [if the masochist in you  wants to get out...

This mesh is properly constructed, thank you, being all quads and  
regular.

Before going to personal attack, you may want to understand properly  
what

is written, before answering aside of the pisspot.

lukep
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://projects.blender.org/pipermail/bf-committers/attachments/20051003/57a1a7c7/attachment.htm