Good call, my bad. We've been working on repairing vtkSurfaceReconstruction for use in our ParaViewGeo product. I accidentally put positive values on the 'negative side' of the surface reconstruction grid, so the contour filter saw two zero contours.
Sorry for the confusion! On Mon, Aug 25, 2008 at 12:48 PM, Moreland, Kenneth <[EMAIL PROTECTED]>wrote: > I don't recall ever seeing anything of that nature with the contour > filter. Perhaps something is going wrong with the surface reconstruction > algorithm. I have not played with it a lot, but the artifacts you describe > sound like those that could be generated if the surface reconstruction does > not generate consistent normals. > > > > It might be helpful to see your data. Is there any way you could post it? > > > > -Ken > > > ------------------------------ > > *From:* [EMAIL PROTECTED] [mailto: > [EMAIL PROTECTED] *On Behalf Of *Chris Cameron > *Sent:* Monday, August 25, 2008 8:49 AM > *To:* [email protected] > *Subject:* [Paraview] vtkContourFilter generates two contours when > requesting only one > > > > Hello, > > I'm using Paraview off cvs somewhere between 3.3 and the latest head. > > I have an Image (Uniform Rectilinear Grid) of 334170 cells and 349600 > points. I have a point Data Array called "Distance" which was obtained using > a surface reconstruction algorithm. I ran the Filters->Common->Contour on > the Image Data and asked for an Isosurface at value 0 for "Distance". > > The output I get is both that contour that looks correct at distance 0, and > a second contour that is decidedly less accurate (very blocky looking) that > 'wraps' the correct contour. In other cases I have had the opposite happen, > that is, the correct contour wraps the less accurate contour. > > Is this the intended output? I expected to only receive one contour if I > only put in one value for Isosurface. > > Thank You, > > -- > Chris Cameron > Mirarco Mining Innovation - e2vo > 705-675-1151 > x5113 > -- Chris Cameron Mirarco Mining Innovation - e2vo 705-675-1151 x5113
_______________________________________________ ParaView mailing list [email protected] http://www.paraview.org/mailman/listinfo/paraview
