@RGcincy I believe that did it! Not perfect. But with a little collision iteration and smoothing, I think that I am on my way!
Thank you for your helpful suggestion.
^__^
Update:
Thought I had it, but apparently, there is still an issue holding the head of the mop together and containing
the strands. There probably has to be a physically weld of the strands to form a band along the root, or like you did with the weightmap on the hanging towel and the clothes pins.
@rakuda did the weight mapping at the head that maikdecker suggested help?
Also, I saw your strands are staying round (which is good!). I've had circular meshes like that collapse into a flat ribbon. Are you changing any of the surface parameters or is it the design of the mesh that is keeping them round?
This is a bit more complex than the bandeau. It has a band at the top to hug the chest above the breasts and a band around the waist too. I used a cylinder but you could use a cone the same way. It took some time to find settings that worked. With default surface properties, the top would slide down and the skirt would rise up and gather around the waist. Feel free to experiment with different property settings.
a. Create a primitive cylinder with the parameters shown in the image. Position the cylinder around a figure with the top just under the armpits (start with a T-pose or with the arms even higher). I also had a slight (-2 degrees) X-rotation to better fit the figure.
b. Follow step 38 c to hide and delete the upper and lower cylinder faces.
c. Follow step 38 e to create 3 additional surfaces: a Top Band using the upper 3-5 rows and a Waist Band of 3-5 rows about where the figure’s waist is located. Select the polygons between the two bands and assign it to a new surface called Top. The skirt portion will retain the name Default.
d. Add a dForce dynamic modifier to the cylinder.
e. We need the Top Band and Waist Band surfaces to cinch quickly to keep the cylinder from sliding down too far. For both of those, I set Contraction-Expansion Ratio to 50%. The Top will also be a bit baggy compared to the skirt, so I set its Contraction-Expansion Ratio to 90% while the Default surface remained at the default 100%.
f. At default density, the bands tended to bunch up, so I set both the Top Band and Waist Band to 500 density. I set the Default (skirt) to 70 while the top remained at the default of 180. Density values below 70 and above 500 caused the mesh to blow up during simulation.
g. Since they need to react quickly, I kept Dynamic Strength for the bands at the default of 1. The Top was set to 0.92 and the Default (skirt) was set to 0.75.
h. Run the simulation using an animated timeline. The image below shows what you get. I could not get the skirt portion to drape down as much as I would like. The tucked waist causes it to go out further from the body and hang as if very stiff fabric. Using a cone instead of a cylinder might help.
i. If you want a shorter dress, select some of the lower polygon rows and hide/delete them using the Geometry Editor. If you simply hide them with the Surfaces Opacity slider, they will still influence the behavior of the lower edge of the cylinder.
This might have already been discussed, but I couldn't find it. What is the setting for solid objects that should ride along with the fabric without deforming?
Examples: Beads in hair, ties on bolos, belt buckles, or (in my case) bullets in a bandolier.
This might have already been discussed, but I couldn't find it. What is the setting for solid objects that should ride along with the fabric without deforming?
Examples: Beads in hair, ties on bolos, belt buckles, or (in my case) bullets in a bandolier.
I don't believe anyone's found a fix for those yet, but generally upping bend resistance and stiffness and reducing simulation strength are your main options.
This might have already been discussed, but I couldn't find it. What is the setting for solid objects that should ride along with the fabric without deforming?
Examples: Beads in hair, ties on bolos, belt buckles, or (in my case) bullets in a bandolier.
I don't believe anyone's found a fix for those yet, but generally upping bend resistance and stiffness and reducing simulation strength are your main options.
A Rigid Follow Node, with the item - as a separate prop or figure - parented to that.
Someone I test for said to turn off the dynamics for things like buttons or ornaments. I'm sure it doesn't work in all instances, but it did work for the coat we were testing so the buttons and belt/buckle didn't slide down,
41. How Resolution Affects Simulation Results. If you plan to design your own props or clothes to use with dForce, it’s important to understand how resolution can influence simulation results. In this section, I use primitive cubes of varying resolution to show what happens.
a. Create a 2-foot cube with 1 division. Create a series of 1-foot cubes with 1, 2, 5, 10, 25, and 50 divisions. Center the smaller cubes slightly above the larger cube. Add a dForce dynamic modifier to all the 1-foot cubes.
b. Hide all but one of the 1-foot cubes and run the simulation. Repeat for all the rest. As you can see in the images below, the more divisions in the cubes the more distortion that occurs under dForce. With 1 division, there is only very slight compression. It takes 5 divisions before you see the sides warping. With 25, the cube is mostly collapsed and with 50 it’s fully collapsed.
1 Division: 2 Divisions:
5 Divisions: 10 Divisions:
25 Divisions: 50 Divisions:
c. Next position a 1-foot cube over each side of the larger base cube. Run the simulation. The lower resolution cubes distort little to none and fall completely off the base cube. Only the 25 and 50 division cubes remain attached as they distort and collapse.
Divisions 1-2-5-10 at Frame 11: Divisions 5-10-25-50 at Frame 11:
Divisions 1-2-5-10 at Frame 18:
d. Now rotate the base cube so the smaller cubes are over a corner. Run the simulation. You get similar results, with divisions 1-2-5-10 falling off while 25 and 50 stick and collapse.
Divisions 1-2-5-10 at Frame 11: Divisions 5-10-25-50 at Frame 11:
Replacing the 1 division base cube with a 5 division base cube had no apparent effect:
dForce cube resolution setup.jpg
471 x 511 - 33K
dForce cube resolution cube 1 frame 11.jpg
466 x 518 - 23K
dForce cube resolution cube 2 frame 11.jpg
461 x 530 - 25K
dForce cube resolution cube 5 frame 11.jpg
481 x 514 - 30K
dForce cube resolution cube 10 frame 11.jpg
441 x 521 - 33K
dForce cube resolution cube 25 frame 11.jpg
466 x 486 - 27K
dForce cube resolution cube 50 frame 11.jpg
481 x 477 - 26K
dForce cube resolution setup side.jpg
382 x 360 - 14K
dForce cube resolution 1-2-5-10 frame 0.jpg
487 x 489 - 29K
dForce cube resolution 1-2-5-10 frame 11.jpg
551 x 532 - 45K
dForce cube resolution 5-10-25-50 frame 11.jpg
538 x 526 - 63K
dForce cube resolution 1-2-5-10 frame 18.jpg
606 x 484 - 46K
dForce cube resolution setup corners.jpg
404 x 400 - 14K
dForce cube resolution base 1 rot 45 5-10-25-50 frame 11.jpg
496 x 536 - 57K
dForce cube resolution base 5 rot 45 5-10-25-50 frame 11.jpg
527 x 536 - 61K
dForce cube resolution base 1 rot 45 1-2-5-10 frame 11.jpg
Thanks for that, very helpful! Does the Daz SubD work for this, or is it only the original base object divisions. (i.e. can you subdivide a cloth using the slider in daz to make it more "flimsy")
SubD doesn't impact the simulation results, only the base resolution matters (unfortunately). SubD will affect the shape of the object, rounding the edges of the cubes.
42. Using Silo 2.5 to Create a Shirt. With the success I had using primitive cylinders and cubes, I thought I’d give a try to making a shirt. Since the basic primitives in Daz Studio don’t have arm extensions, I used Silo 2.5 to create a model. I haven’t used Silo much so my attempt is crude (as you will see) but surprisingly it looks pretty good after applying dForce.
a. This is the model made in Silo. I started with a 9-division per side cube and did some loop splitting, arm extrusion, vertex moving, and face deletion. Not much to look at.
b. One attractive feature of Silo is its subdivision abilities. This is the same crude model after multiple applications of subD. You can see how the square openings have become rounded without any work on my part. I made this more detailed mesh into the control mesh (so it will be preserved upon export) and created an .obj file.
c. Here’s the model after importing the .obj file into DS and placing it around a G3F figure. As you can see, it has poor fit and I had to tilt the head to avoid polygon clinging during the simulation. That wasn’t a problem as I did an animated simulation and moved the head later.
d. Here’s after dForce simulation. Looks good I think!
Hey, this is great to know. I also have Silo 2.5 and I don't really use it because I find the interface is not intuitive. People may dog Hexagon, but it has a simple straight forward UI and is easy to use. I also used to like C4D for the same nice UL.
Thank you so much for sharing your work on this. Much appreciated, and the shirt is just as good as stuff you buy. Nice draping.
Hey, this is great to know. I also have Silo 2.5 and I don't really use it because I find the interface is not intuitive. People may dog Hexagon, but it has a simple straight forward UI and is easy to use. I also used to like C4D for the same nice UL.
Yes, SIlo is quite obtuse. Took me a while to recall how to do anything. I do like the SubD capability as you don't have to model as much. I also have Hexagon but it crashes a lot for me.
Okay, a question... probably there's a real easy solution, but I'm a newb, so...
After using a primitive to shove a part of some clothing around I would like to add the result to the piece of clothing as a morph, so I could "dial it in" when needed. Is there a way to do it? And if there is, is there also a simple, newbie understandable, explanation of "how to do that"?
Okay, a question... probably there's a real easy solution, but I'm a newb, so...
After using a primitive to shove a part of some clothing around I would like to add the result to the piece of clothing as a morph, so I could "dial it in" when needed. Is there a way to do it? And if there is, is there also a simple, newbie understandable, explanation of "how to do that"?
Hide everything but the clothing, set the clothing to base resolution if it has SubD applied (Mesh Resolution in the Parametrs pane), then export as OBJ (noting the export preset used0. Then clear the dForce simulation (button in Simulation Settings), with the to-be-morphed item selected Edit>Figure>Morph Loader Pro, click the Select fiels and pick the exported OBJ, make sure the preset matches the ne from exporting, if it's fitted clothing with a pose applied click the triangle by the OBJ name, find Reverse Deformations, right-click on No and set it to Yes, click Accept.
43. Another Shirt Made with Silo. Since my first attempt was a reasonable success, I tried a second. Instead of using Silo’s built in female figure as a model, I exported a G3M figure as an .obj and imported that into Silo to build the shirt around. I started with the same basic model I did in section 42, but added more polygons and moved a lot of vertices around to better follow the shoulders and arms of the male. I also learned how to rearrange the UV so the texture would follow the shape. Below is the base model, the sub divided model, the shirt before simulation, and the resulting render.
Base Model in Silo:
SubD Model in Silo:
Model Imported into Daz Studio:
Final Render After Simulation:
If it is of sufficient interest, I could put together a tutorial on how to use Silo to create a shirt like this.
EDIT: I've also attached the shirt .obj file. Import it onto a G3M in default pose. You may need to expand the shirt slightly (I used 104% scale, 106.5% z-scale, and 0.87 z-translate) to better overlap the pants. Add the dForce dynamic modifier, add a new pose at frame 15, and run an animated dForce simulation.
Last night, I fired up Hexagon for the first time in years to try my hand at making primitive dforce clothing like RGcincy's. (result: terrible low-res tunic for G3M) It went okay, but although it would dforce the resulting primitive, it wouldn't let me adjust the dforce properties in the surfaces tab until I did the UV-Mapping in Hexagon. (Which makes sense when you stop and think about it, but of course I didn't). So, keep that in mind if you try that out.
Next Goal: Dforceable Pirate Shirt with Frilly Cuffs! I'm thinking that if I model everything right, set up surfaces (Hexagon calls them shading domains) at the wrists and waist, maybe one for the yoke of the shirt, and drop expand/contract percentage a fair amount on wrists and waist and loosen it everywhere else, I can get dforce to do the rest of my modeling for me! :)
@Odaa, glad you gave it a try. The issue about UV's didn't show up for me as Silo must automatically create a UV map, even though it's totally distorted until you remap it yourself. Still have to figure our how to adjust the front and back UV's of the shirt so the stripes match up.
Does Hexagon do subdivision? On Silo, you have to bake the subdivsion before exporting otherwise the saved. obj reverts to the base model.
I also tested a different pose by starting with the default T-pose, moving to the running pose, then lifting the left arm onto the back of the head. The shirt followed along. It is not conformed and has no bones.
A squatting pose did cause the shirt to separate under the left arm and have some polygon distortion on a lower fold. I guess I made it a size too small! So dForce can't overcome everything. I'll have to learn how to rig clothing to see if that helps (barbult posted a link a week or so ago).
Comments
@RGcincy I believe that did it! Not perfect. But with a little collision iteration and smoothing, I think that I am on my way!
Thank you for your helpful suggestion.
^__^
Update:
Thought I had it, but apparently, there is still an issue holding the head of the mop together and containing
the strands. There probably has to be a physically weld of the strands to form a band along the root, or like you did with the weightmap on the hanging towel and the clothes pins.
Weightmapping the part(s) of the strands that meet the head should resolve that problem.
@rakuda did the weight mapping at the head that maikdecker suggested help?
Also, I saw your strands are staying round (which is good!). I've had circular meshes like that collapse into a flat ribbon. Are you changing any of the surface parameters or is it the design of the mesh that is keeping them round?
40. Create a Dress with Gathered Waist.
This is a bit more complex than the bandeau. It has a band at the top to hug the chest above the breasts and a band around the waist too. I used a cylinder but you could use a cone the same way. It took some time to find settings that worked. With default surface properties, the top would slide down and the skirt would rise up and gather around the waist. Feel free to experiment with different property settings.
a. Create a primitive cylinder with the parameters shown in the image. Position the cylinder around a figure with the top just under the armpits (start with a T-pose or with the arms even higher). I also had a slight (-2 degrees) X-rotation to better fit the figure.
b. Follow step 38 c to hide and delete the upper and lower cylinder faces.
c. Follow step 38 e to create 3 additional surfaces: a Top Band using the upper 3-5 rows and a Waist Band of 3-5 rows about where the figure’s waist is located. Select the polygons between the two bands and assign it to a new surface called Top. The skirt portion will retain the name Default.
d. Add a dForce dynamic modifier to the cylinder.
e. We need the Top Band and Waist Band surfaces to cinch quickly to keep the cylinder from sliding down too far. For both of those, I set Contraction-Expansion Ratio to 50%. The Top will also be a bit baggy compared to the skirt, so I set its Contraction-Expansion Ratio to 90% while the Default surface remained at the default 100%.
f. At default density, the bands tended to bunch up, so I set both the Top Band and Waist Band to 500 density. I set the Default (skirt) to 70 while the top remained at the default of 180. Density values below 70 and above 500 caused the mesh to blow up during simulation.
g. Since they need to react quickly, I kept Dynamic Strength for the bands at the default of 1. The Top was set to 0.92 and the Default (skirt) was set to 0.75.
h. Run the simulation using an animated timeline. The image below shows what you get. I could not get the skirt portion to drape down as much as I would like. The tucked waist causes it to go out further from the body and hang as if very stiff fabric. Using a cone instead of a cylinder might help.
i. If you want a shorter dress, select some of the lower polygon rows and hide/delete them using the Geometry Editor. If you simply hide them with the Surfaces Opacity slider, they will still influence the behavior of the lower edge of the cylinder.
Thanks again Rich. This is amazing!
This might have already been discussed, but I couldn't find it. What is the setting for solid objects that should ride along with the fabric without deforming?
Examples: Beads in hair, ties on bolos, belt buckles, or (in my case) bullets in a bandolier.
I don't believe anyone's found a fix for those yet, but generally upping bend resistance and stiffness and reducing simulation strength are your main options.
A Rigid Follow Node, with the item - as a separate prop or figure - parented to that.
Yup, as weird as I still think that is the Ridgid Follow Node works like a charm!
What's a rigid follow node?
https://www.daz3d.com/forums/discussion/203481/setting-up-items-with-rigid-follow-nodes-tutorial
Someone I test for said to turn off the dynamics for things like buttons or ornaments. I'm sure it doesn't work in all instances, but it did work for the coat we were testing so the buttons and belt/buckle didn't slide down,
[oops, posted to the wrong thread....]
41. How Resolution Affects Simulation Results. If you plan to design your own props or clothes to use with dForce, it’s important to understand how resolution can influence simulation results. In this section, I use primitive cubes of varying resolution to show what happens.
a. Create a 2-foot cube with 1 division. Create a series of 1-foot cubes with 1, 2, 5, 10, 25, and 50 divisions. Center the smaller cubes slightly above the larger cube. Add a dForce dynamic modifier to all the 1-foot cubes.
b. Hide all but one of the 1-foot cubes and run the simulation. Repeat for all the rest. As you can see in the images below, the more divisions in the cubes the more distortion that occurs under dForce. With 1 division, there is only very slight compression. It takes 5 divisions before you see the sides warping. With 25, the cube is mostly collapsed and with 50 it’s fully collapsed.
1 Division: 2 Divisions:
5 Divisions: 10 Divisions:
25 Divisions: 50 Divisions:
c. Next position a 1-foot cube over each side of the larger base cube. Run the simulation. The lower resolution cubes distort little to none and fall completely off the base cube. Only the 25 and 50 division cubes remain attached as they distort and collapse.
Divisions 1-2-5-10 at Frame 11: Divisions 5-10-25-50 at Frame 11:
Divisions 1-2-5-10 at Frame 18:
d. Now rotate the base cube so the smaller cubes are over a corner. Run the simulation. You get similar results, with divisions 1-2-5-10 falling off while 25 and 50 stick and collapse.
Divisions 1-2-5-10 at Frame 11: Divisions 5-10-25-50 at Frame 11:
Replacing the 1 division base cube with a 5 division base cube had no apparent effect:
Thanks for that, very helpful! Does the Daz SubD work for this, or is it only the original base object divisions. (i.e. can you subdivide a cloth using the slider in daz to make it more "flimsy")
SubD doesn't impact the simulation results, only the base resolution matters (unfortunately). SubD will affect the shape of the object, rounding the edges of the cubes.
OK, thanks, I was just setting it up to test, so I appreciate you clarifying that!
42. Using Silo 2.5 to Create a Shirt. With the success I had using primitive cylinders and cubes, I thought I’d give a try to making a shirt. Since the basic primitives in Daz Studio don’t have arm extensions, I used Silo 2.5 to create a model. I haven’t used Silo much so my attempt is crude (as you will see) but surprisingly it looks pretty good after applying dForce.
a. This is the model made in Silo. I started with a 9-division per side cube and did some loop splitting, arm extrusion, vertex moving, and face deletion. Not much to look at.
b. One attractive feature of Silo is its subdivision abilities. This is the same crude model after multiple applications of subD. You can see how the square openings have become rounded without any work on my part. I made this more detailed mesh into the control mesh (so it will be preserved upon export) and created an .obj file.
c. Here’s the model after importing the .obj file into DS and placing it around a G3F figure. As you can see, it has poor fit and I had to tilt the head to avoid polygon clinging during the simulation. That wasn’t a problem as I did an animated simulation and moved the head later.
d. Here’s after dForce simulation. Looks good I think!
Hey, this is great to know. I also have Silo 2.5 and I don't really use it because I find the interface is not intuitive. People may dog Hexagon, but it has a simple straight forward UI and is easy to use. I also used to like C4D for the same nice UL.
Thank you so much for sharing your work on this. Much appreciated, and the shirt is just as good as stuff you buy. Nice draping.
That blouse drape (from what is basically a box) is amazing! Lovely job.
Thanks sapat and MelanieL!
Yes, SIlo is quite obtuse. Took me a while to recall how to do anything. I do like the SubD capability as you don't have to model as much. I also have Hexagon but it crashes a lot for me.
Okay, a question... probably there's a real easy solution, but I'm a newb, so...
After using a primitive to shove a part of some clothing around I would like to add the result to the piece of clothing as a morph, so I could "dial it in" when needed. Is there a way to do it? And if there is, is there also a simple, newbie understandable, explanation of "how to do that"?
https://sickleyield.deviantart.com/journal/Tutorial-Requested-Loading-Morphs-in-DAZ-Studio-450007378
Hide everything but the clothing, set the clothing to base resolution if it has SubD applied (Mesh Resolution in the Parametrs pane), then export as OBJ (noting the export preset used0. Then clear the dForce simulation (button in Simulation Settings), with the to-be-morphed item selected Edit>Figure>Morph Loader Pro, click the Select fiels and pick the exported OBJ, make sure the preset matches the ne from exporting, if it's fitted clothing with a pose applied click the triangle by the OBJ name, find Reverse Deformations, right-click on No and set it to Yes, click Accept.
Silo's interface is fully customisable BTW. You can download a few premade UI presets and see if you find a layout you like.
http://nevercenter.com/silo3d/wiki/index.php?title=Customizations_And_Interfaces
43. Another Shirt Made with Silo. Since my first attempt was a reasonable success, I tried a second. Instead of using Silo’s built in female figure as a model, I exported a G3M figure as an .obj and imported that into Silo to build the shirt around. I started with the same basic model I did in section 42, but added more polygons and moved a lot of vertices around to better follow the shoulders and arms of the male. I also learned how to rearrange the UV so the texture would follow the shape. Below is the base model, the sub divided model, the shirt before simulation, and the resulting render.
Base Model in Silo:
SubD Model in Silo:
Model Imported into Daz Studio:
Final Render After Simulation:
If it is of sufficient interest, I could put together a tutorial on how to use Silo to create a shirt like this.
EDIT: I've also attached the shirt .obj file. Import it onto a G3M in default pose. You may need to expand the shirt slightly (I used 104% scale, 106.5% z-scale, and 0.87 z-translate) to better overlap the pants. Add the dForce dynamic modifier, add a new pose at frame 15, and run an animated dForce simulation.
I don't own Silo, but all tutorials for modeling clothing are welcome.
Thanks a lot. You really are the DAZ Guru *bows*
Last night, I fired up Hexagon for the first time in years to try my hand at making primitive dforce clothing like RGcincy's. (result: terrible low-res tunic for G3M) It went okay, but although it would dforce the resulting primitive, it wouldn't let me adjust the dforce properties in the surfaces tab until I did the UV-Mapping in Hexagon. (Which makes sense when you stop and think about it, but of course I didn't). So, keep that in mind if you try that out.
Next Goal: Dforceable Pirate Shirt with Frilly Cuffs! I'm thinking that if I model everything right, set up surfaces (Hexagon calls them shading domains) at the wrists and waist, maybe one for the yoke of the shirt, and drop expand/contract percentage a fair amount on wrists and waist and loosen it everywhere else, I can get dforce to do the rest of my modeling for me! :)
@Odaa, glad you gave it a try. The issue about UV's didn't show up for me as Silo must automatically create a UV map, even though it's totally distorted until you remap it yourself. Still have to figure our how to adjust the front and back UV's of the shirt so the stripes match up.
Does Hexagon do subdivision? On Silo, you have to bake the subdivsion before exporting otherwise the saved. obj reverts to the base model.
I also tested a different pose by starting with the default T-pose, moving to the running pose, then lifting the left arm onto the back of the head. The shirt followed along. It is not conformed and has no bones.
A squatting pose did cause the shirt to separate under the left arm and have some polygon distortion on a lower fold. I guess I made it a size too small! So dForce can't overcome everything. I'll have to learn how to rig clothing to see if that helps (barbult posted a link a week or so ago).