[Sidefx-houdini-list] Skin SOP Flipping
dan.seddon at framestore-cfc.com
Sat Jan 7 10:40:29 EST 2006
What you have is almost there - you have two problems. The first is that
your cookie is making more than one intersection on some frames. To
solve this you want to throw away small curves - use the measure sop to
calculate the size of the curves coming out of merge1, then a delete sop
with "Delete by Expression" and set the expression to "$PERIMETER<5".
This should solve this problem.
Then - it just looks like the point sorting isn't working. Try this:
Create yourself a grid which you are going to use as a "knife" (reduce
it's rows and colums to 2x2 and make it a mesh) - position the grid as
though it were to cut the curve (but only once for each curve - not all
the way through).
Now do just that - create a "Curvesect Sop" and put the output of the
delete sop in to the first input and the grid in to the second. This
should add another point to each of the curves, but in a similar place
on each one - this will become the first and last vertex for each
curve (as all the curves are rebuilt when you do this), so you can now
use another sort sop (sorting by vertex order) to reorder your curves so
they all line up. You shouldn't need the other sort.
Hope that makes sense.
> Rick Walia wrote:
>> ken wrote:
>>> if yes, does their primitive ordering(number) remain consistent from
>>> frame to frame?
> Correction. Mistake on my part. The primitives are changing order. I
> did try sorting the prims by proximity to pt but that didn't help much
> I have uploaded the sample file here:
> I have also posted to the sesi forums here:
> Sidefx-houdini-list mailing list
> Sidefx-houdini-list at sidefx.com
More information about the Sidefx-houdini-list