 | | Re: Why Type 5 Lines? —Don Heyse
|
| | | (...) Tell more about this. I just tweaked a version of ldview to display type 5 lines, but only for surfaces that haven't been replaced by quad strips. It looks good with primitive substitution turned off, but with this automagic wireframing of (...) (23 years ago, 17-Apr-02, to lugnet.cad.dev)
|
| | | |
| | | | | Re: Why Type 5 Lines? —Travis Cobbs
|
| | | | (...) I'm curious to. Everything I know about how OpenGL works says that drawing the strip as a wireframe after drawing it solidly would either produce Z Buffer noise, no wireframe at all, or wireframe all over. However, it may be that there is some (...) (23 years ago, 17-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Kyle McDonald
|
| | | | (...) Well, I believe that there is something called a 'Z offset' that can be used not to actually move the geometry, but to make sure the lines appear 'on top' of the faces. Before I go more into this I need to back up a bit. I currently use the (...) (23 years ago, 17-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Travis Cobbs
|
| | | | | (...) OpenGL has glPolygonOffset, which does what you describe by allowing you to move the polygons further away in the Z direction, where the amount they move can depend on the slope of the polygon in the Z direction. This works great, and I use it (...) (23 years ago, 17-Apr-02, to lugnet.cad.dev)
|
| | | | | |
| | | | | | Re: Why Type 5 Lines? —Don Heyse
|
| | | | (...) That sounds like glPolygonOffset which would be part of the solution because it allows the edges to look clean. We all use that already. (...) I can see how you might accomplish some of that by manipulating glEdgeFlag when you define your (...) (23 years ago, 18-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Don Heyse
|
| | | | | (...) Yeah, the stencil buffer, that's gotta be it. If you've got BFC working then you can use the stencil buffer to count how many times an edge is draw in GL_LINE mode. Edges on back facing triangles won't be drawn, so you'll only count one edge (...) (23 years ago, 18-Apr-02, to lugnet.cad.dev)
|
| | | | | |
| | | | | | Re: Why Type 5 Lines? —Kyle McDonald
|
| | | | I was reading through this more, and while I still don't know exactly how it all works, I think I stumbled onto something. > I dont know how many of you have come across this technique before. I > remember finding an OpenGL demo that did this kind (...) (23 years ago, 18-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Steve Bliss
|
| | | | (...) One problem: transparent surfaces would show the wireframe backside. :\ Steve (23 years ago, 18-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Travis Cobbs
|
| | | | (...) Actually, transparent surface should work fine. As long as they are set to write to the Z buffer when drawn, and their outlines are drawn with Z buffer testing enabled, they should work fine. The wireframe parts that aren't supposed to be (...) (23 years ago, 18-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Steve Bliss
|
| | | | (...) The backside wireframe elements *will* be drawn, that's the point of what Kyle was talking about. The idea is to draw the wireframe of the backside surfaces and it will all be hidden by the frontside surfaces, *except* for the lines around the (...) (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Travis Cobbs
|
| | | | (...) I understand what you are saying, but it won't happen that way if Z buffer writing is enabled during the drawing of the transparent surfaces. The transparent surfaces aren't really transparent. They're just blended with whatever is behind (...) (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Steve Bliss
|
| | | | | (...) Cool. Thanks for taking the time to explain it to me. I didn't realize this about opengl. Steve "please speak slowly and clearly" Bliss (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | | |
| | | | | | Re: Alternate Type 5 line algorithm and other BFC questions... —Kyle McDonald
|
| | | | | [snip] > > Normally you don't bother to enable Z buffer writing when drawing the > transparent surfaces, because they are the last thing drawn, and it won't > matter. However, it doesn't hurt anything to enable it, other than slowing > it down a (...) (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | | |
| | | | | | | Re: Alternate Type 5 line algorithm and other BFC questions... —Travis Cobbs
|
| | | | | | (...) Just check to see if the determinant of the matrix is negative. If it is, the matrix involves mirroring. --Travis Cobbs (tcobbs@REMOVE.halibut.com) (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | | | |
| | | | | | Re: Why Type 5 Lines? —Ross Crawford
|
| | | | (...) Thanks for that explanation, however I think we still need to include type 5 lines in parts, so that programs which don't implement this technique (eg ldraw) can still display the curved edges correctly. Programs which do implement it can (...) (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | |
| | | | | | Re: Why Type 5 Lines? —Travis Cobbs
|
| | | | (...) Sorry, I never intended to imply that it should replace type 5 lines. In addition to the problem of backwards compatibility, this will only work for BFC-certified parts. --Travis Cobbs (tcobbs@REMOVE.halibut.com) (23 years ago, 19-Apr-02, to lugnet.cad.dev)
|
| | | | |
