[Sidefx-houdini-list] wires dynamic problem
guillaume at alamaison.fr
Sat Jan 17 12:50:23 EST 2009
Thanks for the advices :).
Yes, using wires attributes by vertices can be very useful. In my project,
however, the default settings works perfectly.
The problem came from the curve animation as I thought that I could set the
animation only on the first point (the pined one).
But in fact, it looks like the solver needs the "root animation" on the
Here is a simple repro scene : http://www.vol2nuit.fr/basic_wire.hipnc
If you set the switch to zero, all the curve is transformed at sop level and
the wire simulation works correctly.
If you set the switch to one, only the first point of the wire is
transformed at sop level. In this scenario, the wire
simulation stretch a lot between points zero and one.
I don't really understand why the solver needs some sop animation on all the
points who are not pined ?
On Sat, Jan 17, 2009 at 10:30 AM, Louis Dunlevy <louisdunlevy at gmail.com>wrote:
> Another thing to bear in mind is the Sop Solver if it helps to set some
> of those wire object point attributes dynamically. Every wire object
> parameter can be a varying point attribute instead as Mark points out.
> An old hack is to use the primitive sop in the Sop Solver to do a final
> correction on the wire length, just store the original length of the
> wire before the sim and scale it appropriately per step in the Sop
> Solver. With the primitive sop $PT is the first point in every prim so
> getting the origin for each wire is easy.
> If you're adding deformations on top of your sim though you may have to
> adjust your velocities accordingly in the sop solver too. Use the change
> in position caused by the sop solver, divide it by the sim timestep and
> add it on top of the original velocity.. or something of that nature!
> There might be new options for that on the sop solver now though..
> guillaume at alamaison.fr wrote:
> > Yes, I just found the solution :).
> > Only the first point of my wire was animated by hand in SOP. I added this
> > animation to all the points instead
> > of just the first one and now it works perfectly !
> > I found the problem because I didn't understand why the wire dynamics
> > worked well on previous work with some fur animations.
> > I realized that each fur guides sent to the DOP was moving from root to
> > points (as animation came from the copy sop
> > template input).
> > In my current work, I was copying the animation of a character hand on
> > first point of a curve/wire, thinking that all the
> > other points will be animated by the wire solver. Good to know (for me at
> > least :) ) !
> > Cheers
> > Guillaume
> > On Thu, 15 Jan 2009 11:09:32 -0800, Mark Story <mstory at xion.org> wrote:
> >> Hmmm ... that's odd. I don't know if it would be a reasonable work
> >> around, but you can assign those attributes as a point attr and dial in
> >> the exact mount of stretching/bending you need along the length of the
> >> wire.
> >> The particular behavior you describe I've never encountered before,
> >> (I've done a lot of work with Wire DOPs) but am interested in what might
> >> be happening ... if you can't solve the problem, please post a hip file
> >> on one of the forums so we can take a look.
> >> Good luck!
> >> Mark
> >>> I'm working on a simple DOP wire simulation. The setup is just one
> >>> (rather long).
> >>> At SOP level, the root of the curve is constrained on a keyframe
> >>> animation.
> >>> In my DOP network, I'm using "Use Deforming Geometry" in the Wire
> > Object.
> >>> I use a SBD Pin Constraint on the root points ( 0-3 ) and then check
> >>> "Constrain To Animation" to keep
> >>> the SOP animation on the 0-3 points.
> >>> My problem is that I can't find a way to avoid some stretching between
> >>> the pin points and the first point moved by the wire solver.
> >>> To limit the stretching, I set a high value on the Linear Spring
> >>> Constant. It works fine for all the points, except
> >>> between the "pin points" and the "solved points".
> >>> Any tips to stop this stretching is welcome.
> >>> Cheers
> >> _______________________________________________
> >> Sidefx-houdini-list mailing list
> >> Sidefx-houdini-list at sidefx.com
> >> https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
> > _______________________________________________
> > Sidefx-houdini-list mailing list
> > Sidefx-houdini-list at sidefx.com
> > https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
> Sidefx-houdini-list mailing list
> Sidefx-houdini-list at sidefx.com
More information about the Sidefx-houdini-list