From: Acosta, Mark W [Magic Earth LLC] (acostmw++at++texaco.com)
Date: 04/24/2000 14:54:04
Marcin,
Initially, I am setting the static bounding sphere to the encompass all
configurations of the geometry under it. But under certain conditions, and
very infrequently, the bounds for the geometry may shrink or expand.
Typically this is triggered by a user loading/deleting some data. I have no
way to predict what the "ultimate" bounds of my geometry will be initially
because it's driven by what the user does. Again, this is not something
that's occuring every frame. It may take minutes to hours for something to
occur that would cause me to update the static bounds for my geometry.
My reason for using the static bounds was to avoid the overhead of
propagating the changes up the scene graph. Since my data can be fairly
dynamic, I thought having a DYNAMIC bounds might carry a fair amount of
overhead. Maybe I'm wrong here.
I still think if I explicitly change a static bounding sphere on a node it
should propagate the change up the scene graph.
Mark Acosta
Magic Earth LLC
-----Original Message-----
From: Marcin Romaszewicz
To: Acosta, Mark W [Magic Earth LLC]
Cc: 'Brian Furtaw '; ''info-performer++at++sgi.com ' '
Sent: 4/24/00 2:41 PM
Subject: RE: static setBound on a pfNode
A static bounding sphere is one which isn't changing. There is no need
to
propagate something which isn't chanhging up the tree, since it would
have
no effect on the parent bounding sphere. This assumption allows
performer
to save some computation. Since you are changing your bounding sphere
dynamically, create a dynamic one on the pfNode so that performer knows
to
propagate the changes up.
If you absoluytely must have a static bounding sphere, make it a size
that
will encompass all possible configurations for the nodes it contains. If
the gometry the sphere contains does not change too much in terms of
bound
size, this will work pretty well.
-- Marcin
On Mon, 24 Apr 2000, Acosta, Mark W [Magic Earth LLC] wrote:
> Hi Brian,
>
> So you mean do something like this?
>
> geode->setBound (bsphere,PFBOUND_DYNAMIC);
> geode->setBound (bsphere,PFBOUND_STATIC);
>
> Is it normal for static bounds to not propagate up the scene graph or
is
> this a "feature"?
>
> Thanks,
>
> Mark Acosta
> Magic Earth LLC
>
> -----Original Message-----
> From: Brian Furtaw
> To: Acosta, Mark W [Magic Earth LLC]
> Cc: 'info-performer++at++sgi.com '
> Sent: 4/24/00 1:29 PM
> Subject: Re: static setBound on a pfNode
>
> Hi Mark,
>
> It looks like you need to set the mode to PFBOUND_DYNAMIC then
setBound
> to a new bsphere to have it propagate up the chain.
>
> Brian
>
> "Acosta, Mark W [Magic Earth LLC]" wrote:
> >
> > Hi everyone,
> >
> > I have several places in my scene where I do a
> setBound(bsphere,
> > PFBOUND_STATIC) on Geodes that contain GeoSets with dynamically
> changing
> > geometry. The first time I do this, everything works fine. If I
change
> the
> > bounding sphere later on (making it much bigger) I've found that
> Performer
> > doesn't propagate the change up the scene graph which screws up my
> picking.
> > I dumped out the scene graph to verify this. This doesn't seem right
> to me.
> > Is there some way to force the recomputation of the bounds in the
> scene
> > graph?
> >
> > Thanks,
> >
> > Mark Acosta
> > Magic Earth LLC
> >
>
-----------------------------------------------------------------------
> > List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> > Submissions: info-performer++at++sgi.com
> > Admin. requests: info-performer-request++at++sgi.com
>
> --
> ----oOOo---- ----oOOo---- ----oOOo---- ----oOOo----
>
> Brian Furtaw (brian++at++sgi.com)
> Graphics Guru Office:(301)572-3293 Fax: (301)572-3280
> 12200-G Plum Orchard Drive
OpenGL/Performer/OpenInventor/ImageVision
>
> Silver Spring, Maryland 20904 Volumizer/Optimizer/React/PCI Device
> Drivers
>
-----------------------------------------------------------------------
> List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
>
This archive was generated by hypermail 2b29 : Mon Apr 24 2000 - 14:55:35 PDT