Well, try contouring :)... Actually, if you have the proper data structures and LUTs this can be fast. I do this anyways for other reasons.
If your geometrical surface is everywhere orthogonal to the viewing direction then the problem you described can happen. Specifically, in the case of parallel projection you would see nothing - at least this is what I have noticed when I plot rings. In the case of perspective projection, you might see at least an inkling of the outline because of foreshortening. For example, the ribbons used by molecular viz. folks (there are some experts here in this Forum), will not appear on the canvas if you view them exactly end-on. A trick that you could use is to plot the polygon outline also by setting the POLYGON_LINE capab. bit on a copy of the geometry. In the case of the ModelClip-generated rings, you can't employ this trick. You may probably just see intersection points of the edges with the clip plane. An alternative is to clip your elements to generate new tris and then plot their outline - but this is a little complicated for casual use. Don't know...may be there are other hardware tricks. I am wondering if the geometry would still be invisible if the viewing direction is rotated by some degrees ? >On Wed, 9 Oct 2002 17:38:27 +0200, Desiree Hilbring <[EMAIL PROTECTED]>KARLSRUHE.DE> >wrote: >> In "hardware", I think you may be able to use 2 appropriately >> oriented parallel ModelClip planes to achieve the effect. This would >> produce a ring actually, but if the distance between the clip planes is >> small enough it will look like a line contour. Haven't tried it, >> but looks like it would work. > >I tried that for creating 2D profiles from terrain shaped hydrogeological >layers. I created two parallel ModelClip planes with a small distance >between them, cutting my 3D scene. The problem is, that your 3D shapes >have no "thickness". So after cutting the 3D scene and moving your >viewplatform for achieving the 2D-view, you would not see a proper circle, >because when you look directly along the normal of your cutting plane you >can not see that part of the circle, because it is "thin". The outline >will have a perspective effect. I can not explain that very well. The >last picture on my homepage shows a profile of some hydrogeological >layers, created with that approach. You can not say for sure if you are >really seeing the entire profile. > >http://www.ipf.uni-karlsruhe.de/Personen/hilbring/hilbring.html > >Desiree > >> >> >> >On Wed, 9 Oct 2002 07:50:48 -0700, Scholl, Ed <[EMAIL PROTECTED]> wrote: >> >> >Hello- >> >Can anyone give me some pointers to draw the 2D outline obtained from the >> >intersection of a cutting plane with my 3D world? i.e. if I specify a >> >cutting plane that intersects with a sphere, I want to display the resulting >> >circle. >> > >> >-Ed >> > >> > ========================================================================== >> >To unsubscribe, send email to [EMAIL PROTECTED] and include in the body >> >of the message "signoff JAVA3D-INTEREST". For general help, send email to >> >[EMAIL PROTECTED] and include in the body of the message "help". >> >> =========================================================================== >> To unsubscribe, send email to [EMAIL PROTECTED] and include in the body >> of the message "signoff JAVA3D-INTEREST". For general help, send email to >> [EMAIL PROTECTED] and include in the body of the message "help". >> > > ========================================================================== >To unsubscribe, send email to [EMAIL PROTECTED] and include in the body >of the message "signoff JAVA3D-INTEREST". For general help, send email to >[EMAIL PROTECTED] and include in the body of the message "help". =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
