[vtkusers] Re: More info on clipping planes

[Will Schroeder]

Hi James-

Sorry for the delay, I'm catching up with some email now.

The clipping planes are controlled because they adversely impact the 
resolution of the z-buffer. If the planes are too far apart then during 
rendering pixels may be lumped in together as being the same "depth". This 
causes jittering and ugly results.

If you are rendering lines with polygonal primitives then you want to set 
the method:

mapper->SetResolveCoincidentTopologyToPolygonOffset()

on any subclass of vtkMapper. (This is a global flag.)

Will



At 09:30 AM 11/26/2002 +0000, James C. Robinson wrote:
>
>Dear All,
>
>I am having (more) trouble visualising a 3D FE mesh. I have it 95% working,
>viewing the mesh as a vtkUnStructuredGrid. However, the problem is that, for
>certain, spatially large, meshes, detail at the centre (they invariably have
>a narrow cylinder running through them) can disappear as I zoom in to look
>at it. This is because the default setting of the vtkInteracter is to
>AutoResetClippingRange and the value of the near plane (in the method that
>carries it out) has to be 10% of the far plane. This means that when viewing
>a scene with some parts pf an object at 1000m from the camera, anything
>closer than 100m to the camera is "clipped".
>
>The choice is to turn this AutoResetClippingRange off and manage the
>clipping planes oneself. I have almost got that to work, but I am finding
>that the rendering of a vtkEdgeFeatures object that I create to view the
>edges of the elements is adversely affected by changing the setting (see
>attached word file). Has anybody any ideas on this?
>
>Finally, my intuition is telling me that if my PC can handle rendering all
>facets in the actor that maybe I should just set the near and far values to
>say 0.1 and 10000 (assuming that my max dimension of the mesh is say 1000)
>and thus all detail would be visible unless I zoom in closer than 0.1 or out
>such that part of the mesh is more than 10000 from the camera. My reasoning
>is that the original concern with clipping planes may have emanated from
>processing worries for very complex (high numbers of facets) on slower PCs.
>Is this sound reasoning?
>
>Thanks for the attention,
>
>Jim
>
>______________________
>
>James C. Robinson, PhD,
>Chartered Engineer,
>Kepler Engineering Software Ltd.,
>42 Rivergrove,
>Glanmire, Co. Cork,
>Eire
>
>Tel:         +353-21-4822028
>Tel:         +353-87-2393010
>Fax:        +353-21-4822721
>E-mail:     j.robinson at kepler.ie
>______________________
>