[vtkusers] Placing volumes exactly adjacent to each other

[Elvis Stansvik]

2016-09-11 17:01 GMT+02:00 David Gobbi <david.gobbi at gmail.com>:

> On Sun, Sep 11, 2016 at 8:55 AM, David Gobbi <david.gobbi at gmail.com>
> wrote:
>
>> On Sun, Sep 11, 2016 at 8:17 AM, Elvis Stansvik <
>> elvis.stansvik at orexplore.com> wrote:
>>>
>>>
>>> I thought that with e.g. a 2x2x2 volume with spacing 1.0 1.0 1.0, I
>>> would get 2x2x2 area occupied when the volume is mapped into the world.
>>>
>>
>> Let's get straight to the point.  In order to seamlessly display two
>> adjacent volumes, the volumes must share a face, e.g. the rightmost face of
>> the first volume must have the same data points as the leftmost face of
>> second volume.  I.e. the data points on the edge must be duplicated between
>> the two volumes.
>>
>> This isn't specific to VTK, or even to volume rendering.  If one is
>> analyzing/visualizing a data set in a manner that requires interpolation,
>> and the data set needs to be broken into pieces, then adjacent pieces must
>> share data points at the edges.
>>
>
>  Actually this isn't specific to interpolation.  Any time the
> analysis/visualization of the pieces requires any kind of neighborhood
> filtering whatsoever, there must be data values shared between adjacent
> pieces.
>

Ah yes, that makes sense.

My confusion in this slightly stream-of-conciousness-style thread (sorry
'bout that!) was that I didn't really grasp how a 2x2x2 source volume could
give rise to a 1x1x1 cube in the world. But it was just that I didn't
understand how rendering works and how the extents of the volume relate to
what I see rendered. I think I understand now.

Regarding ensuring the volumes are seamless: So far, the visual result in
this particular case seems acceptable enough to me, without having to worry
about "stiching" them together by making them share values at the boundary.
But I'll keep it in mind and do some more testing, thanks for bringing it
up.

Elvis


>
>  - David
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtkusers/attachments/20160911/a3cea2ad/attachment-0001.html>