topic Re: nCloth Attract to matching Mesh and Input Attract in Maya Dynamics Forum

nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Fri, 11 Aug 2017 05:58:41 GMT

I am simulating an nCloth Jacket but there are areas that I don't want to deform.

I have connected the outMesh of the skinned jacket geo to the inmesh of the cloth shape and set the input attract to 1.5

I have painted vertex maps for some attributes;

deformability is black

rigidity is white

input attract is white

in the area that i don't want to deform but the cloth still deforms

I added an Attract to Matching Mesh constraint and painted the strength to 1 in the same area

but the cloth still deforms.

Is there a way to actually make the cloth selectively deform in certain areas and not in others?

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Fri, 11 Aug 2017 11:55:29 GMT

Hi!

The paint values are no absolute values they are multiplicators.

Paint values will be multiplied with the corresponding attribute values in nClothShape.

If the values in nClothShape are "0", then painting has no effect.

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Fri, 11 Aug 2017 12:16:11 GMT

thanks dude, that part is understood in theory but doesn't seem to work in practicality.

Like an input attract value of 1.5 being multiplied by 1 (white) should result in no simulation, 100% (or 150%) input/skinned mesh should be the final result.

Shouldn't the attract to matching mesh at a strength of 1 override all of those settings anyway? Are they not designed to work together?

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Fri, 11 Aug 2017 12:38:23 GMT

Hi!

Works here (Maya 2017 update 4).

Please also try with a simple plane, maybe there is a problem with your object or setup.

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Sat, 12 Aug 2017 03:47:23 GMT

Unfortunately as with a lot of stuff (Phoenix comes to mind) they ONLY work on the simplest of setups, the minute you introduce complexity the whole system falls apart.

What is the pink locator and why is the vert next to it that is fully black not acting like cloth?

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Sat, 12 Aug 2017 08:03:41 GMT

Attract to Matching Mesh

Not working at all

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Sat, 12 Aug 2017 09:20:55 GMT

OK - so it seems that ALL settings are irrelevant in the face of rigid body collisions. Essentially the body inside the jacket is colliding with the jacket and pretty much just destroying any attribute or constraint.

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Sat, 12 Aug 2017 12:58:33 GMT

Hi!

"What is the pink locator and why is the vert next to it that is fully black not acting like cloth?"

In my example i just pinned 3 vertices for testing the simulation.

If you want a soft body (that's what nCloth is) to behave like a rigid body you would need infinite high values for resistance and Rigidity.

So set the values as high as you need and increase the value for substeps at the nucleus solver.

Though when i look at your example i would choose a more manual controlled approach (maybe by using nConstraints).

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Sun, 13 Aug 2017 01:57:29 GMT

yeah exactly why I was trying to use Attract to Matching Mesh (AtMM) - because the input attract attribute doesn't work as expected.

However the AtMM constraint deforms on collision with rigid bodies - I suspect all the others will as well - so when the cloth is constantly colliding with the rigid body shoulders it gradually gets pulled further and further out of shape and by the end of the sim it is a mess.

Like why wouldn't you allow the user to treat certain areas of the cloth like rigid bodies (or no sim skinned mesh) by setting the input attract to 1???? It is a perfectly simple solution - piping the skinned mesh into the cloth input - if it worked. instead you have to resort to wrapping your final object to four or five different objects with different cloth settings, it is so damned primitive!

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Mon, 14 Aug 2017 16:29:11 GMT

Hi!

It looks like you just have just chosen the wrong setup / nConstraint for your animation.

For example "Point to Surface" works very well here. There should be an easy solution for your project, just use a different setup. (You should not need more than 1 additional object to drive the simulation).

Without knowing more about your scene i cannot help you further (but I also don't do tutorials here in the forum or help making a project)

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Mon, 14 Aug 2017 22:02:14 GMT

Riiiiight, so because Input attract doesn't work I have chosen the wrong setup???? Never mind the fact that it doesn't work, or that i've read about this set up working in other situations again it must be me... no possible other solution.

No one is asking for a tutorial or for help with a project dude

Has anybody told you the language you choose is quite insulting?

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Mon, 14 Aug 2017 22:05:40 GMT

No one is asking for a tutorial or for help making a project dude.

Has anybody told you the language you choose is quite insulting?

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Tue, 15 Aug 2017 04:10:27 GMT

Hi!

Input attract does work here. I also tested the tools like painting maps and nContraints and they work fine for me (Maya 2017 update 4).

But as i said without knowing more about your setup i can't say why it's not working for you or you don't get the result you want. My assumption is that you use a tool in a way it wasn't designed for. It's not my intention to insult people, but to offer a solution or workaround to solve a problem.

Because the tools you named work here without errors my suggestion still is to try a different setup.

Testing an other setup is a very general recommendation but the only one i can make in this case.

If you think there is a bug or a tool not working as it should, please upload a scene-file or fill a bug report.

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Wed, 16 Aug 2017 03:04:12 GMT

I guess it depends on what your definition of "working" is as to whether it works or not - for me it definitely doesn't work the way I would expect it to work and I have read numerous posts experiencing the same difficulty. I'm sure it works fine in isolation but an isolated plane simply does not approximate my set up. If you dropped that plane over a rigid body and had a skinned, animated mesh as the input it might start to not "work" the way it doesn't work in my set up.

What I have found though is that because input attract doesn't work I have to add a constraint over the top to essentially do exactly what input atttract should be doing. And thus why I felt the need to post this on a forum because why have input attract at all if you can't set absolute values or you have to apply constraints which do the same thing only better. it is either redundant or broken.

The point on surface constraint works much better than the Attract to Matching Mesh - if the target is the rigid body then the points in the constraint set are totally rigid, but if I target the skinned geo it does have a little flex in there which is perfect.

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Wed, 16 Aug 2017 14:08:34 GMT

Hi!

The nCloth itself is a soft-body and all internal calculations are based on this. Input attract is only an attract, not an override, it still behaves to the internal rules of a nCloth and reacts to other objects part of the simulation based on set and calculated influence values, except if you change Input Attract Method to "Lock values ...".

For a point on surface constraint an initial Rest Length is set and kept (by default, but you can change this), so if the object displaces, this has an effect on the connected nCloth Components. If you really want to lock vertices to an object and introduce 100% stiffness, set Constraint Method to "Weld", for example for a cloth partially stapled onto a piece of wood.

This mimics the behavior of real world cloth where you also need some additional objects combined with the cloth to introduce stiffness.

I can't see any problem that cannot be solved by a maximum of one intermediate object to drive the simulation.

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Thu, 17 Aug 2017 01:34:55 GMT

Yeah these constraints are working like a charm. Set up and working on one shot then load new animation - this is the first frame of sim.

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Thu, 17 Aug 2017 03:22:35 GMT

nCloth - the gift that keeps on giving.

After setting the rest position to avoid the above self destruction from the constraints I converted a wrapped low poly geo of the pants to a rigid body.

The collision display is of the highlighted pants rigid body object.

Awesome

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Thu, 17 Aug 2017 05:50:18 GMT

Hi!

"this is the first frame of sim"

This looks indeed really bad. It's the same scene/character just an other animation clip, keys,...?

"The collision display is of the highlighted pants rigid body object."

You created a Passive Collider and the collision detection is so far off?

Do you have a nCloth in the simulation, i have just seen in Maya 2018 the collision preview only updates with a nCloth in the scene.

Are you able to upload one of these scene-files or an other one for testing?

Re: nCloth Attract to matching Mesh and Input Attract

magillaGuerilla — Thu, 17 Aug 2017 05:58:51 GMT

sorry, I can't upload. The jacket sim was resolved by setting rest shape to start shape on the frame that the nucleus becomes active. I haven't used this before, but it works so... problem solved

The rigid body collision for the pants only did this when referenced. I had to remove it from the referenced file and recreate it in the scene that it was referenced into - if that makes sense. Then it worked exactly as expected. Previous rigid body stuff has been fine when referenced including the upper body collider which is still referenced and working correctly, I haven't been able to determine why this instance of the pants collider was not working.

Re: nCloth Attract to matching Mesh and Input Attract

mspeer — Thu, 17 Aug 2017 06:25:48 GMT

Hi!

Thanks for the answer.

"I haven't been able to determine why this instance of the pants collider was not working."

No idea. I have never referenced a simulation, but i will test this.