Whatever this is, it's just insidious. I have models that are reading fine for days then suddenly out of nowhere it is corrupted upon opening a saved scene. It's almost as if the corruption is occurring upon the previous scene save. I can go back to the scene save prior and compare the edges or vertices or poly that are reported bad and in the prior save they are different components than the ones being reported corrupt in the current scene. At least they are in different places on the geometry.
A lot of these models do have user normal clusters but that's all as far as clusters go. I guess I'll try to hunt them all down and delete them. The worst offenders weren't created in SI and I have no pedigree on what they were made in but it's clear that SI can't handle them safely. -- Joey Ponthieux LaRC Information Technology Enhanced Services (LITES) MYMIC Technical Services NASA Langley Research Center __________________________________________________ Opinions stated here-in are strictly those of the author and do not represent the opinions of NASA or any other party. > -----Original Message----- > From: [email protected] [mailto:softimage- > [email protected]] On Behalf Of Matt Lind > Sent: Monday, July 06, 2015 3:22 PM > To: [email protected] > Subject: RE: ERROR : 2000 - Mesh structure error (corruption) > > You are experiencing exactly what I experienced at Carbine that prevented > upgrade from Softimage 7.5 for 5+ years. If you can tell me with certainty > which version of Softimage you made the geometry, I may be able to help > you salvage some stuff. > > I reported a ton of corruption issues from XSI v6.x thru 2012 SP2. All those > versions have hidden gremlins like what you just uncorked. The specific > gremlin and it's solution varies depending on the version of the software the > geometry was created and/or last modified prior to being opened to 2013 or > later. Geometry created in 2013 SP1 or later should not have any of these > issues. > > One thing you could check is to see if your geometry has polygon clusters. > If so, try rebuilding them, then deleting the original polygon cluster(s). > I suggest doing while geometry operator loading/evaluation is disabled in > your user preferences. That tended to fix a lot of our problems. > Triangulating geometry also helped. > > I also suggest you fix the issues in the version of Softimage which they were > created and not try to fix it in 2014 or later as the fixes to those problems > actually prevent you from fixing them in the latter versions. For example, if > the scene was created in Softimage 2011, then fix the problems in Softimage > 2011. > > > Matt > > > > Date: Mon, 6 Jul 2015 13:51:34 +0000 > From: "Ponthieux, Joseph G. (LARC-E1A)[LITES]" <[email protected]> > Subject: RE: ERROR : 2000 - Mesh structure error (corruption) > To: "[email protected]" > > It would seem that this is mostly geometry related. It also seems that it is > an > age issue as I have some Models that were created in Soft about 5 years ago > and created no problems then but are demonstrating issues now. What's > more bizarre is that one mesh has topology errors that I am certain did not > exist back then. It literally boils down to a simple 64x64 square grid of > quads > that was generated from a loft and later converted to a mesh. There is an > area of about 15 or so polys that are mangled, missing, and non-manifold, > kind of like you see in some DWG to mesh conversions. But there is > absolutely no apparent reason for it. I have other geometry that came over > from Maya that was originally NURBS and a lot of older Viewpoint geometry > as well. I see the problem a lot with those models. But that isn't what > surprises me. It's the custom modeling created in Maya and Softimage about > 5 years ago that really doesn't make sense. It's like the entire mesh database > for any geometry that exhibits the problem is unstable. The minute you > delete offending points, edges or polys the problem moves to a different > part of the mesh. Its as if the geometry, when read by SI, was read wrong > and mangled. You have to delete the entire mesh and recreate it. Just as you > said, it's a lost cause. The only thing with these models that seems to be > consistent is their age and the fact that they started existence as a NURBS > model and at some point converted to mesh. > > -- > Joey Ponthieux > LaRC Information Technology Enhanced Services (LITES) MYMIC Technical > Services NASA Langley Research Center > __________________________________________________ > Opinions stated here-in are strictly those of the author and do not represent > the opinions of NASA or any other party.
