I've had the same problem in the past and my solution, purely on the
modelling side, was to offset my glow polygon along the axes that points
towards the eye (+Y I think). Then the billboard is positioned at the light
source but the glow is drawn a little closer towards the eye. So as you look
more obliquely, the glow polygon is not immediately clipped in half. It does
start to get clipped at some angle but the effect is much less serious. For
me it was adequate for all my small lights in the scene and you might need to
experiment with the offset distance which depnds on the light size. I used
the occlusion method too but only for really bright objects like the sun with
lens flare etc.

Regards, Simon
_______________________________________________________________________

Simon Mills
Modelling & Simulation Section (TEC-SWM)        Tel: +31 (0)71 565 3725
European Space Agency (ESA/ESTEC)               Fax: +31 (0)71 565 5420
Postbus 299, 2200AG Noordwijk               e-mail: [EMAIL PROTECTED]
The Netherlands                       http://simulation.esa.int
_______________________________________________________________________


                                                                             
             Roderick Kennedy                                                
             <Roderick.Kennedy@                                              
             evos.net>                                                    To 
             Sent by:                   osg users                            
             osg-users-bounces@         <[email protected]>       
             openscenegraph.net                                           cc 
                                                                             
                                                                     Subject 
             04/01/2007 17:26           RE: [osg-users] Light point depth    
                                                                             
                                                                             
             Please respond to                                               
                 osg users                                                   
             <[EMAIL PROTECTED]                                              
                negraph.net>                                                 
                                                                             
                                                                             




No now I think about it, that wouldn't work. What you need is for the actual
depth test to depend on the sprite centre, which can't be done with the usual
z-test. Sounds like the occlusion test is the way forward.

From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Roderick Kennedy
Sent: 04 January 2007 15:57
To: osg users
Subject: RE: [osg-users] Light point depth

How about a billboard sprite shader that takes z depth from the sprite
centre?

From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Zach Deedler
Sent: 04 January 2007 15:37
To: 'osg users'
Subject: RE: [osg-users] Light point depth

Hmmm, looks like there is no simple solution.  For now, I'll just try to
offset the light points more, and put the more elegant solutions on the back
burner.  Thanks for all the ideas everyone.



Zach




From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Paul Martz
Sent: Thursday, January 04, 2007 10:17
To: 'osg users'
Subject: RE: [osg-users] Light point depth


 I don't care if the light point occludes objects within a few meters.  Would
 using a PolygonOffset work with a meter offset?
I believe PolygonOffset won't work for you when viewing the light at an
obtuse angle, say from the side, and slightly in front of, the vehicle. In
such a case, the further headlight would need a greater depth offset than the
near headlight to in order to completely avoid intersecting the vehicle body.

In general, whether PolygonOffset produces acceptable results or not varies
depending on available depth buffer precision, and how you've set your
near/far planes.

Paul Martz
Skew Matrix Software LLC
http://www.skew-matrix.com
303 859 9466_______________________________________________
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/

Reply via email to