We had tons of corruption problems in earlier versions of Softimage. In many cases the geometry was fine, but there were bad clusters which operators needed to read to do their work. When the operators received bad data, the errors would be raised. Those problems could be solved by deleting and manually rebuilding the clusters.
We also experienced a synchronization issue between the geometry’s data structure and the viewports. In some cases using editors such as the texture editor would display corruption plain as day with mangled UVs and all, but the viewports would show the geometry perfectly intact. Freezing the construction history often remedied the problem, but only temporarily. We often had to export to dotXSI or obj to cleanse the object. Importing geometry from another software usually results on some type of cluster information being generated whether it be a sample cluster for user normals, or a polygon cluster for a local material. When we had to push scenes from current versions of Softimage to previous versions, we tried using .dotXSI format and found it created bad polygon clusters. The workaround was to triangulate the geometry in the export to .dotXSI. Not sure if it applies to your situation. Matt From: [email protected] [mailto:[email protected]] On Behalf Of Ponthieux, Joseph G. (LARC-E1A)[LITES] Sent: Thursday, June 19, 2014 1:11 PM To: [email protected]; [email protected] Subject: RE: geometry corruption Detect corrupted polygon meshes and clusters upon load/freeze was on. If I manually select two of the known offending points and perform info selection I get this SelectGeometryComponents("B787_Mrg.B787_Mrg.pnt[278]"); AddToSelection("B787_Mrg.B787_Mrg.pnt[276]", null, true); // ERROR : 2000 - Argument 1 (Proxied) is invalid // ERROR : 2000 - Argument 0 (InputObj) is invalid // ERROR : 2001-EnumElements - Argument 0 is invalid InfoSelection("B787_Mrg.B787_Mrg.pnt[276,278]"); But I get this for any selection on the model. I’ll leave the error check off and see if that helps. Thanks -- 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. From: [email protected]<mailto:[email protected]> [mailto:[email protected]] On Behalf Of [email protected]<mailto:[email protected]> Sent: Thursday, June 19, 2014 3:59 PM To: [email protected]<mailto:[email protected]> Subject: Re: geometry corruption Hi Joey, is the "detect corrupted polygon meshes and clusters upon load/frezze" (PREFERENCES DATA MANAGMENT) on , if on switch to ignore/of, also what happens when filter points (info selection) number of points same or less?!, still the error on load? This is I would first check, good luck Walter "Ponthieux, Joseph G. (LARC-E1A)[LITES]" <[email protected]<mailto:[email protected]>> hat am 19. Juni 2014 um 21:33 geschrieben: Howdy, I have a model which is throwing intermittent errors upon the scene being read. The error says: // ERROR : 2000 - Mesh structure error (corruption) was detected on mesh _______ while loading it. See below for precise information about corrupted components. // ERROR : 2000 - List of corrupted vertices: 276, 278, 4535, 11632, 14802, 14865, 14880, 14892, 14898, 14944 // ERROR : 2000 - List of corrupted edges: 523, 36828, 40348, 40447, 40529 So what’s weird about this? The mesh is about 60K polys and about five edges and ten vertices(all associated to the 5 edges) are being reported as corrupted. There are nearly 70 copies of the SAME model in the scene. Generally only one or two throw the error during a scene read and I have seen about 6 or 7 different ones throw the error during the lifetime of this project. The geometry was originally created by me in Maya as NURBS and exported to Soft. There is no external geometry not created by me. I have the offending geometry isolated as a model, I can read it(import) into a clean scene with no problems or errors. However, upon importing it, during interrogation of the offending geometry I find that: 1. It is a single mesh, though there are islands within the mesh. 2. There are no non-manifold edges or vertices associated to the offending components. 3. There are no duplicate edges or vertices associated or near the offending components. 4. There is nothing unusual about the listed edges or vertices, they all look clean, adjoined, and normal. 5. A couple of the edges are border edges, but they also appear clean. 6. Selecting and modifying these vertices and edges have caused several crashes. 7. These crashes during interrogation are occurring on a model that was created in Soft immediately after Maya export. As a result, the errors appear to have nothing to do with the master scene that this model was imported into afterwards. Generally I’ve been ignoring the errors for the duration of the project, until recently. If the scene loads everything it seems to hold up just fine regardless that it generates the error warnings, but I’m starting to run into issues where the scene crashes upon loading once it spits out the error on the geometry. Subsequent reads sometimes succeed and some fail. I’m getting concerned that with time the scene may eventually fail upon read indefinitely. I’d really rather not drag this back to Maya for a cleanup if I can avoid it, I’ve applied clusters and local materials to the mesh from within Soft. If anyone has any suggestions for a “cleanup” procedure, I’d love to hear them. Thanks -- 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.
