detail 0 (Collision-1) bounds exceed that of shape
by Marton Gyorgy · in Torque 3D Professional · 05/10/2011 (3:17 am) · 5 replies
Hi,
I got the following warning, but my object does not register collision on its sides, just on its top. It's a cube. I exported it from Milkshape, with the following:
Box0 - the actual cube
Collision-1 - duplicated cube to use as collision mesh
Bounds0 - somewhat bigger bounding box. I am sure it includes the above meshes.
Could somebody tell me what's wrong?
I got the following warning, but my object does not register collision on its sides, just on its top. It's a cube. I exported it from Milkshape, with the following:
Box0 - the actual cube
Collision-1 - duplicated cube to use as collision mesh
Bounds0 - somewhat bigger bounding box. I am sure it includes the above meshes.
Could somebody tell me what's wrong?
About the author
#2
However now it works with a Collision-# named mesh and a Collision-# named node! Strange.
05/10/2011 (6:42 am)
I tried with Col-1 but it was not recognized as collision mesh. By the way the milkshape exporter documentation says collision-# named meshes will be used as collision meshes but the DTS node list on TDN says they should be named Col-# and Collision-# are just markers.However now it works with a Collision-# named mesh and a Collision-# named node! Strange.
#3
Yes...when you export a mesh grouping in Ms3d that has all vertices weighted to a single bone[or 'no' bone at all, [iirc], you'll get a node of the same name[and this is 'good', helps engine determine it's an object and you get proper engine behavior]. It sounds as though your first attempts, the Col-1 node was not created...and perhaps led to your collision issue.
When you export a mesh grouping as a 'skinned mesh'[multiple bone influencing a single vertex] you do not get this node creation, since the multi weighted mesh is already consider a node/object[for exportation schema]. This may have happened to your shape, although I would think the naming convention would create the object properly....??
To really get a definitive answer to what is happening to the art, you'll have to share it to another artist[like myself or others]. It is really difficult to diagnose a problem without seeing/werking all the pieces...sounds like you got it sorted with the second naming...
What version of exporter and Ms3d are you utilizing? That helps in the 'diagnosis' as well...
Did you bind your 'collision' mesh to more than one 'bone', originally???
What build of the T3D engine are you using...things are in 'flux' until the Preview becomes final, and you may be using a earlier 'beta' version...with some 'unknown' issue??
good luck!
the takeaway: I think something got 'goofy' in your Ms3d setup??
05/10/2011 (8:05 am)
ah, 'Strange'/not so much[lol]...Yes...when you export a mesh grouping in Ms3d that has all vertices weighted to a single bone[or 'no' bone at all, [iirc], you'll get a node of the same name[and this is 'good', helps engine determine it's an object and you get proper engine behavior]. It sounds as though your first attempts, the Col-1 node was not created...and perhaps led to your collision issue.
When you export a mesh grouping as a 'skinned mesh'[multiple bone influencing a single vertex] you do not get this node creation, since the multi weighted mesh is already consider a node/object[for exportation schema]. This may have happened to your shape, although I would think the naming convention would create the object properly....??
To really get a definitive answer to what is happening to the art, you'll have to share it to another artist[like myself or others]. It is really difficult to diagnose a problem without seeing/werking all the pieces...sounds like you got it sorted with the second naming...
What version of exporter and Ms3d are you utilizing? That helps in the 'diagnosis' as well...
Did you bind your 'collision' mesh to more than one 'bone', originally???
What build of the T3D engine are you using...things are in 'flux' until the Preview becomes final, and you may be using a earlier 'beta' version...with some 'unknown' issue??
good luck!
the takeaway: I think something got 'goofy' in your Ms3d setup??
#4
I got 'collision' mesh/object by naming my Mesh Grouping: "Collision-1"
and I also got 'LOScol' mesh/object by naming my Mesh Grouping: "LOSCol-9"
I hope this helps?! Line Of Site collision meshes need to have their appending numbers start at '9' for them to be considered a 'los' object by the engine[at least thru ms3d/dtsplus!], naming your LOS anywhere below '9' will make them pure 'collision' and not noticed by projectiles and raycasts.
The export dialog seems to do some goofy things if I try to change the LOD value for a mesh object named collision...?? It always fills the entry field with '0' after I leave the mesh edit section??
I name my mesh makers in the Ms3d GUI, not really in the mesh edit section of exporter; unless I'm doing some autoDetail LOD's...
I did see the meshes in the ShowToolPro, but I haven't dropped them into a mission to see if they're collide-able....
05/10/2011 (12:30 pm)
I have a working example now, Marton. I got both types of 'collision' meshes exporting and may have noticed an oddity with the dialog of the exporter.I got 'collision' mesh/object by naming my Mesh Grouping: "Collision-1"
and I also got 'LOScol' mesh/object by naming my Mesh Grouping: "LOSCol-9"
I hope this helps?! Line Of Site collision meshes need to have their appending numbers start at '9' for them to be considered a 'los' object by the engine[at least thru ms3d/dtsplus!], naming your LOS anywhere below '9' will make them pure 'collision' and not noticed by projectiles and raycasts.
The export dialog seems to do some goofy things if I try to change the LOD value for a mesh object named collision...?? It always fills the entry field with '0' after I leave the mesh edit section??
I name my mesh makers in the Ms3d GUI, not really in the mesh edit section of exporter; unless I'm doing some autoDetail LOD's...
I did see the meshes in the ShowToolPro, but I haven't dropped them into a mission to see if they're collide-able....
#5
05/21/2011 (3:41 am)
Thank you very much for your detailed answer I am afraid I will have to return to it in the future. The collision prolem is solved now but I am not sure how. I think there were other things not working as expected and I did not correcty identified the problem. 
Torque Owner Rex
BrokeAss Games
Cheers!