Hi Terry,
I don't know if this helps, but the way I do quad tree paged terrain
databases is to set up the scene graph like:
PagedLOD
first child (lowest res one) is subgraph will real
geometry/textures in place
second child is a not loaded by default, but loaded via file right
The external file then contains:
Group
PagedLOD (as per above for bottom left corner)
PagedLOD (as per above for bottom right corner)
PagedLOD (as per above for top right corner)
PagedLOD (as per above for top left corner)
The database recurses on for ever :-)
Robert.
On 12/19/06, Terry Welsh <[EMAIL PROTECTED]> wrote:
Thanks Murray. I'll be sure you get you a copy when the loader is
ready for stress-testing.
Now I have a little logic question for anyone who has done terrain
paging. I want terrain tiles at level 1 to be replaced by multiple
tiles level 2 simultaneously. But if each tile has its own LOD center
then level 2 tiles pop in at different times as you approach them, and
this causes holes to appear in the terrain.
I can fix this problem between levels 1 and 2 by setting the same
center on all these tiles. However, this center is inappropriate for
the level 3 children of a level 2 tile. What I really need is for
each tile to have have a separate switch-in center and switch-out
center.
My scene graph is set up in units that look like this. A PagedLOD has
multiple child PagedLODs and one filename representing the tile at
that level. Each of those child PagedLODs contains multiple PagedLODs
and a filename just like this one.
PagedLOD --- PagedLOD
PagedLOD
PagedLOD
PagedLOD
FileName
Does this explanation of my problem make sense? Is there a better way
to set up the graph to solve this problem?
- Terry
> Hi Terry,
>
> I'd be interested in supporting your effort to create a (partial) MetaFlight
> loader in any way that I can. My lab uses both OSG and Vega Prime (w/
> LADBM) on a regular basis and would welcome the addition of this capability
> to OSG. I can't promise months of programming support, but we have several
> bright programming sparks in the lab and would be happy to support testing,
> etc., if it would be of help.
>
> Cheers,
>
> Murray
>
_______________________________________________
osg-users mailing list
[email protected]
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
_______________________________________________
osg-users mailing list
[email protected]
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
