I could do that, but why bother if it works ok with 5 drape frames and normal simulation?

Try it out, with the dress parented, and then the figure's body scaled down a little. 

Quote - > Quote - Doesn't parenting the prop to the figure before the simulation negate any benefits of shrinking or growing the figure because the prop will shrink or grow with the parent? I know I've had that happen to me before. I've always waited to parent the finished simulation to the figure.

No, it won't negate the effects. In fact it creates a very smooth stretch. When the cloth is "clothified" it seems to stay the size that it starts with on the first frame of the simulation, and is then forced to stretch if the figure grows, or to drape if the figure shrinks.

I think it depends on which part of the figure you parent it to because when I parent a dress to the "BODY" and then do scaling, the dress scales with the figure. 

Quote -
I think it depends on which part of the figure you parent it to because when I parent a dress to the "BODY" and then do scaling, the dress scales with the figure. 

Really? From my experience, parenting a dress to the body and scaling the figure down in the first frame will make the figure stretch into it, and if you scale it up it will make it looser. You need to make sure that the figure is 100% in size in the target frame. I can show you examples if you want, but try it for yourself, it's always worked for me.

Note that the waistband stays the same way due to that it's constrained.

Anyways, there you go :)

That it's too buggy.

That it's not worth the time when most of the times it'll give you artifacts and blotches and holes.

That it's annoying to make the animation just for that.

And so on~~

I take it back. I take it ALL back. Look at this. It's completely stunning, and I'm taken and stunned and saving to buy a commercial MarvelousDesigner license so I can make products like this and better like crazy. I don't want to sound arrogant, but this image made my jaw drop, and it's all because of dynamic cloth. I'm in love.

@Grappo
Wow, that is a simple yet impressive result. Who would have thought......

@Afrodite-Ohki
Very nice drape. I also like the apparent thickness of the material. Is that displacement?

Quote - @Afrodite-Ohki
Very nice drape. I also like the apparent thickness of the material. Is that displacement?

Only on the collar and sleeve straps... that was a cheat. Those parts (that were modelled as trims above the cloth) had major pokethroughs after the simulation, so I put a 4mm pure-white displacement on them LOL

Hi Maira,

Smart, and looks convincing to me!

Thanks! :D

Awesome thread!!

I've wanted to do a feudal Japan comic but all the flowing clothing can only be done really with dynamic cloth. Lots of great techniques here.

Quote - Frame 30. Body Scaled to 143%,, ZScaled to 158%. Dress parented to body. Notice the dress has scaled exactly with the body! This happens consistantly when I parent to the figure with ALL props as it is supposed to!  Otherwise you would always have to rescale each parented prop individually! This is why I don't parent dynamic clothing to the body until after the simulation is finished if I'm shrink wrapping.

Have you scaled it in frame 30?

You need to scale it on frame 1, and reset all values to 100% on frame 30 in order to get the effect you see on the picture I posted.

And it doesn't look like you've run the simulation on your pictures. The point is to do the following to get different 'size' for the dress, tighter or looser.

1. Parent the dress to the figure.

2. Scale the figure up and down as you please on frame 1. (The dress should follow)

3. Make sure that the scale of the figure is set to default 100% on the target frame, (ie frame 30)

4. Clothify the prop in the clothroom, set the simulation length to at least the target frame.

5. Run the simulation. Different scale for the figure on frame 1 will give different draping results on frame 30. Perfect to scale down x and z if you want a tighter dress.

Here Is the dress NOT parented to the body. Started with the figure at 100% on frame 1 and ended with the figure ZScaled to 150 at frame 30. As the figure grows into the dress it effectively shrink wraps. If it had been parented to the figure like my first example it wouldn't have shrink wrapped as the figure scaled out.

Somebody else in the thread has already pointed out to you that you are doing the exact opposite of what I do to get the same effect. Regardless, props scale with the figure when you parent them.

Quote - And it doesn't look like you've run the simulation on your pictures. The point is to do the following to get different 'size' for the dress, tighter or looser.

1. Parent the dress to the figure.

2. Scale the figure up and down as you please on frame 1. (The dress should follow) < This is exactly what I said about parented props. They scale with the figure and do NOT shrinkWrap to the figure!!

3. Make sure that the scale of the figure is set to default 100% on the target frame, (ie frame 30)

4. Clothify the prop in the clothroom, set the simulation length to at least the target frame.

5. Run the simulation. Different scale for the figure on frame 1 will give different draping results on frame 30. Perfect to scale down x and z if you want a tighter dress.

You are NOT shrink wrapping. You are merely scaling the figure to different scales and the dress follows the scale! Exactly what I said it would do if parented to the body. This is NOT the shrink wrap process!!

In order to Shrink Wrap, either the dress must be shrunken to the figure or the figure must "GROW" into the dress and stretch it out to the figures new scale.

I get your point, and I see what you are doing. However, I cannot see that the prop scales with the figure if it's parented. If it did, I would not get the results that I showed earlier in the post. 

Normally it would scale, but it seems that the cloth simulation "locks" the size of the prop when it starts it. Unless the -prop- has been given any scaling values by itself, which it hasn't.

Having to make the figure larger than normal in the target frame would be counter productive to me if I am just trying to get a tight or loose fit for the clothing. If you look at the picture I posted you can clearly see the effect.

The dress is not scaling with the figure, but is being stretched out as the figure grows into it. It wouldn't do this if it scaled with the figure.

Nope. I parent dress to figure. Scale figure in frame 1 (and making sure last frame has scaling set back to 100%), then I run the simulation. Have been doing it that way all the time, with tons of different clothes.

I am wondering though, in your example, is your figure's zscale 100% on frame 30?

Ok, I see what is happening now. When using your method, and you start the simulation, even though the dress scaled down with the body before the simulation, it snaps back to it's original scale at frame 1 during the drapping process while the body stays scaled down. This is why I have been unable to shrink wrap a dress down to a figure while parented because if I scaled the dress up at frame 1 in order to shrink it down to fit by frame 30 it just snaps back to it's default scale when the simulation starts.

Yeah it seems you have been doing it the other way around. If you are using the method I'm using, and you want to shrink wrap the dress, you need to scale the figure down rather than up at frame 1.

I used to do it without parenting, but then I had to scale the dress manually with the figure in frame 1, which never got really exact, and constrained groups would get wonky, with pokethroughs or gaps.

Looks nice. :) What matters is that we get the results we're after.

Quote - Looks nice. :) What matters is that we get the results we're after.

Actually, as part of this particular discussion. Accurate descriptive information on the process matters very much. Otherwise people will not be able to use this thread to recreate the same effects you display.

Quote - The girl in the image in very big in the shoulders, breast, hips and thighs.  She's also taller than average, a feat I accomplished by slightly increasing the length of her legs and abdomen.  As has been suggested in this thread, I thought I'd shrink her down in frame one and allow her to "grow" into her actual size.  It didn't really work that well, though. Most of the time, cloth simulations work for me without too much effort.  Occasionally I'll have problems, but the vast majority of my dynamic simulations progress without any trouble at all.  I've scratched my head for a long time at people who find it so confusing.

My initial experiences with dynamic cloth have been quite the opposite - they have nothing but massive (confused) effort with rubbish as result. Which is why I threw together that little "Try-This" - hope it makes sense to the newbies - a basic starting point, which is what I needed, and which I'm pretty sure more dynamic neophytes need. I've since changed parameters a fair bit with better results, but I still think that those starting points will get even the most frustrated wanna-do-dynamics started with 99% of what's out there.

I only do dynamic cloth, now. And I have KobaltKween to thank for that brilliant recipe!

One simple question, as the content creator that I am: for dynamic cloth, quads or tris for the mesh?

From my observation, and from arguments here, it doesn't seem to matter. What does help is higher polygon count. It allows for smaller bends.

I demonstrated why quads are trouble quite clearly I thought in the last dynamic cloth thread.

Link to said thread please, BB?

Attached Link: Dynamic cloth - the cloth room For Compleat Dummies - page 7

If you look on that page 7 a little earlier than the link, you'll see I posted a link to some cloth simulation test squares I built with various polygon shapes and densities. You can use these yourselves to test settings.

Thank you muchly mister :3 Goes to read some

In spite of this, for some clothes, quads ended up with nicer drapes for me than tris did. Sometimes tris look better. Note that these were straight triangles. I haven't worked much with uneven tris because Wings doesn't have much support for it.

What is really important to me, possibly most important is how the clothing is modelled, and where. For collars, seams etc, you want to give them surfaces that outline their shape, which will enable the clothing to drape more realistically. It also makes it much easier to texture the UV map later. Say you have a shirt, you don't want tris or quads crossing over where the seams go. Make sure the seams go where they should, so they can bend properly there. :)

So, there is no "best-mesh-type" for all types of cloth, sounds like. All depends what the mesh is meant to deform to / look like when deformed / be UV-map-friendly / other considerations that haven't been explored yet.

This has got to be the most exciting areas of Poser to get into, and possibly one of the least charted: dynamic cloth making. It'll be a while before I venture there! Gonna mess with a few more conforming -> dynamic conversions first: it's a real eye-opener!

Quote - So, there is no "best-mesh-type" for all types of cloth, sounds like.

Correct.  That goes for wrinkles, too.  And the underlying mesh handles different mesh type cloth differently.

it's bunching too much.  :huh:  i'm using wings, too.

a different set of problems with it, when i try to make it conforming.

To deal with the bunching I usually make a hires version of the cloth for final renders. How many polys are those shorts? Fabric in real life bunches too, and a high poly count tends to make it more forgiving. 

Sometimes adding more calculation steps per frame can decrease bunching. 

the poly count was close to 3,000. 

i left the calc steps at 2.0. 

thanks!  - i'll experiment again with the steps

Quote - i made this a few days ago, before reading quads vs tris. 

it's bunching too much.  :huh:  i'm using wings, too.

a different set of problems with it, when i try to make it conforming.

You should also see what your cloth density settings are set as. Lowering them should help with the bunching as well.

If it needs to be good, i.e for a commissioned illustration, I export it back to my modeller and subdivide and smooth polygons until it's right. I am not so worried about polygon count...

I know Sculptris and Marvelous Designer create delaunay tris when modeling, but I don't think they can convert quads to delaunay tris. Any ideas on an affordable program...?

Quote - the poly count was close to 3,000. 

i left the calc steps at 2.0. 

thanks!  - i'll experiment again with the steps

For a final version you could easily subdivide the shorts. I love detailed drapes, and for that you need more details in the mesh. It'll take longer, but looks very nice. 10k polys shouldn't be a problem for the shorts.

My free Mystic Shirt has a highres version with around 30k polys. It takes a while longer to calculate and is a case of more extreme high detailed mesh, but gives really detailed folds.

http://www.renderosity.com/mod/freestuff/details.php?item_id=62376

Quote - > Quote - the poly count was close to 3,000. 

i left the calc steps at 2.0. 

thanks!  - i'll experiment again with the steps

For a final version you could easily subdivide the shorts. I love detailed drapes, and for that you need more details in the mesh. It'll take longer, but looks very nice. 10k polys shouldn't be a problem for the shorts.

I've found that if i want a bit more detail in a given area, I'll go ahead and add some mesh (subdivide) that area only - did this for this dress:

...right around the abdominal area, where I knew I needed a bit more mesh to make finer folds. Is this a reasonable strategy? No clue. This is a conforming dress that was well-enough designed to convert nicely to dynamic, with a few tiny mods.

Quote - > Quote - > Quote - the poly count was close to 3,000. 

i left the calc steps at 2.0. 

thanks!  - i'll experiment again with the steps

For a final version you could easily subdivide the shorts. I love detailed drapes, and for that you need more details in the mesh. It'll take longer, but looks very nice. 10k polys shouldn't be a problem for the shorts.

I've found that if i want a bit more detail in a given area, I'll go ahead and add some mesh (subdivide) that area only - did this for this dress:

...right around the abdominal area, where I knew I needed a bit more mesh to make finer folds. Is this a reasonable strategy? No clue. This is a conforming dress that was well-enough designed to convert nicely to dynamic, with a few tiny mods.

oooo.  i've been wanting an off one shoulder dress for Olivia.

i have one in work for Antonia, Olivia and Victoria4, currently i finish the antonia version and the olivia version must be ok for the end of this week and the V4 version a few days later.

i don't understand what a hex triangle is? 

Hm... I wonder how I get C4d or Zbrush to convert shapes into that.