i very often have to apply a mesh from a 3d-surface to its flattened 2d-surface. this worked great in rhino 2, but after being forced to abandon the AGlib kernel the “applymesh” never worked good for me and until the current rhino version i realised no improvement. i attached two examples:
in the first V6 fails applying, V2 gives a good result.
in the other examples V6 also fails, V2 comes close, but is also unusable.
i have tons of other examples giving me lots of hardaches during the past years i also tried ApplyMeshUVN without any luck, maybe i´m missing something here.
my questions are:
will we ever get a reliable applymesh again?
can anybody help me with a workaround / solution for examples 2-4 as they are part of a running job?
Hello - thanks, I’ll take a look. @brt1 - it may be that you are using UnrollSrf to get the flats, correct? You should use UnrollSrfUV instead to retain the UV space on the flattened object. The changes to UnrollSrf in V5 made the result more accurate and predictable but did not maintain the UV - the old version was reinstated as UnrollSrfUV for cases like this.
hm - looks like that is not it… how are you generating the meshes?
If I get meshes from the surface, it seems to work with either unroll.
i used unrollsrfUV (i usually use this, i have better experiences with it). my plugin offers two possibilities (smash and unrollsrfUV)
btw. in this case, i finally solved the described issues with smash.
the surface is part of a polysurface (all surfaces shrunk, as many untrimmed surfaces as possible). it is important to get watertight meshes. if i mesh the single surface, the meshing will be simpler and work better for applying. but the result useless (not watertight)
sorry for the delay, i was on an offline holyday for a week.
i tried your UnpackTextures advice, unfortunately no reasonable output.
i attach a file with the results from the following workflow:
polysurface with unpacked textures
flattened (smashed / unrollsrfUV)
i also made a comparison with rhino 2 (AgLib-kernel) output. i exported the flattened V6 result and ran the applymesh in V2 on both flattened variants. V2 gives much better (still not perfect) results on both, the unroledlsrfUV and smashed surfaces: