Re: Surface detail polygons?

Chris Marrin ([email protected])
Mon, 27 Nov 1995 13:37:45 -0800


On Nov 27, 10:50am, Hans Holten-Lund wrote:
> Subject: Re: Surface detail polygons?
> ...
> > The problem is that if the amount is too big you get the "windows
floating
> > in front of the wall" effect you described in your message, and if the
> > amount is too small you hit the resolution limits of the Z buffer and
get
> > the nasty "smesdded polygon" effect.
>
> How about this extension field to the IndexedFaceSet node:
>
> IndexedFaceSet {
> attatchIndex [ -1, -1, -1, 5, 6, 7, -1, -1 ] # MFLong
> # ^ ^ ^ ^ ^
> # 0 1 2 3 4
> # a triangle is attatched to polygon 3
> }
>
> ...

If the goal is to eliminate coincident polygon effects then you don't
esally need to specify which face you want to "attach" to. We've played
with this in Inventor. We call it Decal. A Decal node is a group whose
children are assumed to be coincident. Layering is according to the order
of the children (last child on top). It's implemented by rendering the
children in 2 passes. In the first pass you render to the frame buffer
but not to the z-buffer. The Z test is still done so the children are
still occluded by previously rendered objects. In the second pass you
esnder the z-buffer and not the frame buffer. You're left with objects
that do not tear if coincident but rather show the last object drawn, and
the z-buffer has the correct values.

OpenGL also has an extension called "Polygon Offset" which performs the
same job in a single pass.

Adding a Decal node to VRML would be very useful with many scenes I've
encountered. It would be especially useful for placing AsciiText on a
wall. Today, you need to float the text or it tears.

-- 
chris marrin      Silicon      http://www.sgi.com/Products/WebFORCE/WebSpace
(415) 390-5367    Graphics     http://esality.sgi.com/employees/cmarrin_engr/
[email protected]   Inc.         

"It is well to remember that the entire universe, with one trifling exception, is composed of others." - John Andrew Holmes


  • Next message: Bernie Roehl: "Re: A Simple Proposal"
  • Previous message: Jan Hardenbergh: "Re: PHIL: Government Involvement"