Do you have many branchgroups? Or are the objects all in one? I think java3d does branchgroup culling by creating the bounding shape of the group and checking whether it is in the current boundingSphere. So one of either might be faster.
I remember there being a posting about a year ago about performance difference between using multiple branchgroups.
Sorry - haven't used Java3D for a while.
Mattijs
Gilson Laurent wrote:
Hello
I'm trying to create a scene with a lot of static, non-moving, non-animated, no-capabilities-set shapes. Most shapes have between 4 and 6 triangles and have been stripped and indexed. To give you some numbers, these are the StripiferStats for the whole thing:
num orig tris: 133944 num orig vertices: 401832 number of strips: 49083 number of vertices: 251386 total tris: 153220 min strip length: 1 max strip length: 48 avg strip length: 3.121651080822281 avg num verts/tri: 1.6406865944393683 total time: 2753 strip length distribution: 1=924 2=30250 3=195 4=770 5=16146 6=320 7=67 8=107 9=44 10-19=224 20-49=36 50-99=0 100-999=0 1000 or more=0
On the screen everything looks all right. The framerate is a bit low, but with so many stuff in the scene, that's OK.
BUT: scene.compile() takes 3-4 min on a XP1700+. It is called only once after creating/adding all shapes.
If i leave out compile() the framerate drops 50%. So what can i do to speed up compile() ?
(And i'm already working on an LOD, but the data is not LOD-friendly)
thx & cu
=========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
=========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
