Ha! a legend, c'mon guys, I'm not dead yet!

Oh, I cannot thank you enough, Gérald, for creating the dynamic draping tool I have been wishing for for YEARS!

By the way, I'm thinking of trying VWD on shoes and boots. Those have always been troublesome when fitting them to figures they weren't created for, and maybe VWD can help solve those problems too, who knows?

The 'nail to collision' and especially the selective rigidity settings just could be perfect for footwear, but I haven't yet seen anybody use VWD for them yet.

Have a wonderful weekend, Cheers!

Erik

Thanks guys : )

There's so much more to try, so little time indeed!

Hi Gérald,

The original OptiTex mesh has 6881 vertices, 13499 faces, the subdivided one has 27196 vertices, and 53990 faces (according to MeshLab).

It was just an example of a garment that in its original form, although created to be draped dynamically in OptiTex, does not drape well in VWD with ANY setting for the cloth parameters.

That is of course not a problem caused by VWD at all, but it is something to check, and if necessary correct, before loading stuff into VWD.

This simple check will save a lot of frustration and wasted time if you happen to catch a 'bad one'.

That's all.

Cheers!

Erik

The pose in the image below may not look like anything special, but that freebee Optitex coat in its original form was impossible to drape that way in VWD.

The collar collapsed, the shoulders crept up to her ears, and the sleeves simply exploded, whatever settings I tried.

And all that time I was struggling without any success, I was just two clicks away from a perfect result: Edit Object - Convert to SubD

Cheers!

Erik

And here is the very same coat, draped in VWD using exactly the same settings, but this time I subdivided it in DAZ Studio before sending it to VWD (Loop type, level one) .

Same VWD settings, same coat, same pose, but giving a completely different result!

ONLY because the mesh of the coat had many more vertices for VWD to play with.

Much better drape, and hardly any errors to repair in ZBrush after draping.

Cheers!

Erik

Here's an example of a mesh that really is too coarse to work nicely in VWD

No matter what settings I tried, the sleeves were too rigid to bend properly. So every time there was one elbow sticking out through the fabric, and the other sleeve kept falling away from the arm completely on every try.

Luckily I could nudge that uncooperative sleeve back over the arm using the (absolutely brilliant) Dynamic draping option, but still, MANY repairs in ZBrush were necessary to make that coat look like this.

That particular coat is one of the older (unrigged) OptiTex freebees, so originally intended for dynamic draping, but clearly in need of subdivision before sending it to VWD!

I would have saved myself a lot of time if I had thought of checking the structure of that mesh earlier.

Cheers!

Erik

You're very very welcome : )

Here's another example of a complex bundle of hoses or cables that was a joy to create using VWD.

Cheers!

Erik

Here's a close up of the mesh structure with two levels of Loop SubDivision in DAZ Studio.

For SubDivision to work without breaking up at the edges, the mesh must be fully welded after thickening, also a trivial task in Rhino.

Cheers!

Erik

Hi Smaker1,

I use Rhino-3D, but might as well use ZBrush or 3dCoat for this.

In Rhino its a convenient two-minute job, even though it isn't a true mesh modeller like the other two.

Here's a screenshot from DAZ Studio, on the left the original OptiTex Doctor Coat, on the right a thickened version (1mm). You can see clearly that the UV's have been retained, inside and out.

You can also split the inner and outer layer, so you can use a different texture for the lining of the garment.

Cheers!

Erik

Here's what you could do with that kind of VWD stuff (just for fun!)

Cheers!

Erik

Thanks Gérald, both for appreciating my experiments AND adding useful features to VWD : )

Of course I will add some of my pictures to the VWD gallery, anything I can do to 'spread the word'.

Here's another example of the method I used for the camera straps.

It can be used for hydraulic or electrical conduits, or cables, or hair ornaments, whatever.

Something like this would have taken me hours to do 'the old way', now using VWD it just takes a few minutes!

Cheers!

Erik

@Smaker1: Did you try that method? You'd first have to get that shirt into the trousers. Getting the legs of a very loose fitting pair of pants into high (rubber) boots would be a similar case. You would want the trousers to 'billow' or bulge over the top of the boots, not stick like cling-film onto the legs of the figure wearing them.

Here's the same girl again, I adjusted the back of the vest and attempted to drape a camera strap in VWD.

Not particularly successful yet, but I think I may know a better way : )

According to Gérald it is impossible for VWD to predict how any individual mesh will respond to the 'cloth' spring parameters, because the vertex density and structure of every model is different.

What VWD would need for this is an analysis module to measure those mesh characteristics, and then calculate a 'normalizing' conversion factor to compensate for these differences.

But that wouldn't work for every model either, because different parts of one single model can have different structures, so the converted settings would still not be correct for all parts of such a model.

The only thing we can do now is study the structure of individual meshes we want to drape ourselves, and learn by trial and error what effect the settings in VWD have on those meshes.

I think a 'catalogue' of images of typical meshes at different densities, patches of, say, 10x10 cm, with appropriate VWD settings for those meshes would probably help shorten that arduous process tremendously.

Erik

Hi Gérald, Thanks for having a look at this.

I think I prefer my method, because there is no way to predict where the edges of the (in this case) pants will be after draping the shirt. So you cannot limit the 'cling film' effect to the shirt vertices that will be covered by the pants, and that, I think, is essential for a natural look.

Here's a render with the shirt compressed by a vest, using the same method of inverting the vest normals for the (static) simulation. I limited the interaction to the parts of the shirt that are covered by the vest by making the rest of the shirt vertices 'fixed' in VWD.

The result is not perfect yet, I still need to adjust the shape of vest for the pose, and I think I'd better do that in ZBrush, but it shows the pushing-in effect.

It occurred to me that from the start you must have only considered working with layers of clothing from the innermost layer stacking outward, and never from an outer layer pushing layers of clothing inwards. But that does happen often enough in real life (as I think my example shows) to make it a relevant work flow for VWD.

Cheers!

Erik

Hi Smaker,

I really was hoping someone would come up with a more elegant method than this, but it seems like there isn't any.

My first step was to make an animated simulation for the trousers on G2F in VWD.

Then I used the Geometry Editor in DAZ Studio to flip the normals of the VWD-trousers (in frame 1 of the animation).

Next step was an animated simulation of G2F + VWD-trousers as collision objects, and the shirt as cloth.

That went well, until the hands came too close to the hips, then the sleeves of the shirt got sucked off the arms by the pants.

So, I changed plan, and did a static T-pose drape of the shirt as cloth and G2F and VWD-trousers (with inverted normals) as collision objects.

With the shirt draped between the trousers and G2F in the T-pose I then did an animated simulation like before, but this time WITHOUT the pants.

For this step I nailed the shirt vertices that I could see were compressed by the pants in the static drape to G2F, so they would stay put during the animated simulation.

After this step I exported both the posed trousers and the shirt to ZBrush for some adjustments and repairs, and then rendered them in Octane (for DAZ Studio).

All in all a wee bit convoluted, so a simpler method would be very welcome!

Erik