Thanks for the response, Gerrit. The trees do have a blend chunk associated
with them, which was added by the VRML loader. 

I'm still a little confused as to why the BB are returning the wrong sort
order. The water (which is shown in the pics) has a center point far below
that of the tree and so should always be deemed behind the tree....Unless
the distance checks are only done in 2D space relative to the cameras
direction, I can't really see why this would occur because in 3D space the
center of the plane is always father than the center of the tree...

I'll probably parse through the scene dissecting each material as a
post-load process and make the trees alpha test only, but I would really
like to find a nice solution for this problem since it will likely show up
again when we start adding particle effects and such.

On another note, do you (or anyone else) have an idea about the dance
frustum? When using frustum culling and other culling technique (e.g.
occlusion culling) I get objects randomly disappearing, even if only for a
frame or two. I'll try to do some more thorough tests, but has anyone seen
this type of behavior before?

Thanks again,

E.

---
Eric Maslowski
Research Computer Specialist
University of Michigan 3D Lab

Autodesk 3D Studio Max Certified Trainer

email:  [EMAIL PROTECTED]
office: 734-615-9699
mobile: 734-730-9904

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Gerrit Voss
Sent: Wednesday, July 11, 2007 02:48
To: [email protected]
Subject: Re: [Opensg-users] Some Very Strange Artifacts as wellas
Transparency Sorting Issue


Hi,

On Tue, 2007-07-10 at 15:26 -0400, Eric Maslowski wrote:
> Hello all,
>   I've read through the archives and documentation, but it seems that
there
> is something inherently wrong with my frustum and transparency sorting.
(see
> image below) Objects that are clearly distant are drawn in front of closer
> objects, but it also depends on orientation of the camera. Looking at the
> pics it shows me that the objects aren't being sorted properly, but their
> bounding volumes are correct. When visualizing the volumes, I also get a
> strange frustum in the distance. 
> 
> Could this frustum actually be the "cameras" frustum which is somehow
> independent of the camera? (see link below for example movie of it's
> behavior) If so, this would explain some other oddities I'm seeing related
> to culling. Also of note is that if I turn off AutoFrustum the dancing
> frustum disappears when drawing volumes, but I get a warning "left & top
> parallel" and the transparency still issue exists.

form looking at the picture the problematic part is the transparent
surface that covers the ground. 

The problem I can see is that the the sorting happens according to the
bbox centers. So the comparison for sorting will compare the distance 

camera <-> center bbox tree

to 

camera <-> center bbox plane

which given the the center bbox plane is behind the camera for the first
two sets of pictures and the tree bbox center is in front of the camera
gives the interesting result that the sorting flips.

Short question do you use transparency or alpha keying for the tree ?.

Alpha keying might help as it would take the tree out of the
transparency (IIRC).

Another option would be to use the material keys to force a fixed
rendering order (also IIRC I would have to check if the backend does
the right thing)

kind regards,
  gerrit







-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users

Reply via email to