|Message: Re: Smart voxelization for parameterised volume||Not Logged In (login)|
Click on the Forum title, e.g. on the "Forums by Category" page, to read a sequence of postings to the Forum and its threads all in one page. If you are only interested in one thread or the thread following a specific posting, click the thread or the posting, which takes you to a smaller page, which contains only the part you are interested in and may be easier to navigate.
Messages are "chained" if there are only replies at the first level, i.e. 1/1.html, 1/1/1.html etc. In case of "chained" messages the message number is replaced by the icon and there is no indentation.
Inline: Display the subject line only or also the text of the posting(s); for the choice "All" the "Outline" choices are switched off.
|1||0||1||no text / full text of posting|
|2||1||All||text for level 1 only / text for All postings|
Outline: Choose the depth of the posting thread, successive toggle controls provide increasing detail.
|1||2||1||2 levels / 1 level (original posting)|
|2||3||2||3 levels / 2 levels|
|3||3||All||3 levels / all levels (all postings)|
Turns out there are two ways to get proper tracking with moving volumes. As John pointed out, the voxel data structure is built at initialization time and does not change during the run. If a volume leaves its original voxel, the tracker loses sight of it. Furthermore, it is not possible to turn off voxelization for parameterised volumes. The two options are therefore:
1) Simulate only one event per run. Provided that all volumes are stationary during that event, the voxel structure that was built at initialization time will remain valid. A new structure is then built for the next run.
2) Call SetSmartless(0) for the mother volume of the parameterised volume. This forces the tracker to use a single, big voxel that contains all the volumes. Since there is only one voxel, no translation or rotation can cause a volume to change voxels.
When using 1), a lot of time is spent building the voxel structure, but the tracking is fast. For 2), little time is spent building the voxel structure but the tracking is slow. For my application, 2) turned out to be faster overall. It's also worth pointing out that 1) does not lend itself to multithreading.
In my original question, I asked why SetSmartless(nDaughters) seemed to work for a ball but not for a ball with a surface layer. The reason was that my test case included only a single ball. In the "core only" case, there was only one volume, so only a single voxel was created even though I set Smartless > 0. In the "core + surface" case, the parameterisation contained two volumes, and setting Smartless > 0 resulted in more than one voxel being created. The tracking then failed because some moving volume changed voxels.