Ok, some light started to shine! Thanks for the suggestions! As you probably can see from http://www.fm.vok.lth.se/Staff/Priv/DM/water-turbine.html the problem still persists. The light that I mentioned was about the fact that I noticed with the connections. If the picture is small they start to mix front with back, but not as much as in the case with isolines. Anyway, when I zoom in, the the triangles start to be cristal clear. Following this, I have zoomed in also the picture with the isolines and the interferences dissapeared. Now my problem is what can I do to have nice isolines? I tested the net on two different machines, one with nvidia graphic card and another with i810 graphic card, and the result was the same. I run linux on both machines but one is debian woody with 2.4.26 kernel and second with slackware 10.0 with 2.6.10 kernel. The option that you mention for the Isosurface control, I did not set, to be onest I did not know there is such an option before. Can you give me other hint to what I can do, or how to present this picture in order to look nice? I tried to put an AutoColor directly after the ShowBoundary, and the result is what you see in the first picture. It is nice and is not mixing again the two surfaces but it is no longer as clear as it would be with the isolines.
Thank you, Dragos On Wed, May 25, 2005 at 12:27:16PM -0400, Chris Pelkie wrote: > Very weird all right. A few more things to try: > > Does this persist if you rotate the object around? If it changes from > opaque to semi-opaque, maybe it's a renderer bug related to your > graphics card (what system are you on?). > > Turn on the AutoAxes grid: I'd be interested to know if the blade is > really semi-transparent or not. I would expect to see white through the > isolines, but I'm seeing a faint blue. If we can't see the axis grid > through the blade, then the blade is there and opaque and we have to > figure out why the isolines show 'through'. > > You might also try this. Turn off the isolines. Do ShowConnections on > the ShowBoundary output (to get the triangles rather than the > connections of the blade's tets) and Collect the ShowConnections and > ShowBoundary objects. This would also be plotting lines on the surface > of the boundary. If the weird display persists, I'd suspect a render > bug on your hardware. If it looks similar to the isolines, I'm not sure > what it would be. > > You haven't somehow applied a really large Options value for attribute > "fuzz" to the isolines, have you? This oddly named attribute is used to > tell the renderer about the Z-buffer depth of objects. If it were set > rather high, the lines would appear to sit on top of the blade when > rendered. This is not normal, so is probably not a likely explanation > here. As far as I know, the only way to add "fuzz" is to explicitly > with Options (or of course through imported data objects). Normally, DX > attempts to render lines coincident with surfaces on "top" of the > surface, but that should only apply to the frontmost isolines in your > case. > > On May 25, 2005, at 10:46 AM, Dragos MOROIANU wrote: > > > > Thank you Chris for your reply, > > The object is a class field containing 3D data with connections of > > type > > tetrahedra. > > The domain consists of a computational volume from which a turbine > > blade was > > extracted (logical substraction operation). In this way, a part of > > the boundary surface that confines the domain represents the > > surface of the blade. > > You can see the domain, the blade and the pressure isolines at > > http://www.fm.vok.lth.se/Staff/Priv/DM/water-turbine.html > > Although the blade seems to be opaque in the first two images, in > > the last one > > you can see the isolines from both sides of the blade. > > I tried in many different ways to elude this problem but with no > > result. If > > you can see the picture and give me a hint to solve it, I will be > > gratefull. > > > > Dragos > > > > > > > _______________________________ > Chris Pelkie > Scientific Visualization Producer > 622 Rhodes Hall, Cornell Theory Center > Ithaca, NY 14853 (607) 254-8794 > > --
pgp33Jk6Rs7XW.pgp
Description: PGP signature
