Hi Robert,

In this case, I'm loading an openflight file that was built with one
vertex per light point node. That's apparently a bad thing to do, but
I don't have control over it.

A minimum bounding sphere would make the problem unlikely to show up.
I think it would have to depend on the light's min pixel size. This
will still leave weird corner cases though. The "correct" thing would
be to dynamically update the bounding sphere as the eyepoint changes,
but I'm guessing the performance hit would be too much.

Max

On Mon, Jan 11, 2010 at 12:43 PM, Robert Osfield
<[email protected]> wrote:
> Hi Max,
>
> I'm not sure that the comment is quite accurate enough.  Small feature
> culling can be used with osgSim::LightPointNode when you have a
> collection of light points which create a large enough bounding sphere
> about them.  The problem arises when you have a small number of light
> points close together or most problematic a single light source on its
> own.  Perhaps having a minimum bounding sphere radius for a light
> point node would be appropriate to get round this.
>
> In your case how many points are you placing in the LightPointNode?
>
> Robert.
>
> On Fri, Jan 8, 2010 at 5:33 PM, Max Bandazian <[email protected]> wrote:
>> Hi,
>>
>> It's my understanding that osgSim::LightPointNode requires that small
>> feature culling be disabled to work correctly, and that this is the
>> intended behavior. I suggest that this be documented in the header,
>> since at the moment the only documentation of this behavior is the
>> mailing list archives. I've attached a version with descriptive
>> comment included.
>>
>> Thanks,
>> Max Bandazian
>> _______________________________________________
>> osg-submissions mailing list
>> [email protected]
>> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
>>
> _______________________________________________
> osg-submissions mailing list
> [email protected]
> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
>
_______________________________________________
osg-submissions mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org

Reply via email to