Dynamic Hair Tutorials using dForce are up
![DrNewcenstein](https://farnsworth-prod.uc.r.appspot.com/forums/uploads/userpics/847/n5MOELM3KA5JO.jpg)
Finally finished my 4-part YT video series on how I use the Geometry Editor to make Dynamic Hair with dForce.
This is not from other sources. This is what I've learned doing it myself. I've done 7 hairs to date, as shown in the image. Yes, that's V3 Donna Hair on the right.
The others are Persephone, Lustrous, Charm, Elegance, Spring, and Dream Hair.
First three videos are up:
Ep 1
Ep 2
Ep3
Episode 4
Total running time is about an hour and a half. Total time spent doing or saying anything you may find useful, roughly 20 minutes.
Post edited by DrNewcenstein on
Comments
Thanks dude, will have a look. Almost done getting 4 uploaded yet?
It's at 86% according to YT's status thingy. This one's 35 minutes, so it's a big 'un.
Ok, cool, part 4 seems to be the most improtant one lol.
Saved the best for last lol
Ok, Part 4 is up now. Updated the original post with the link.
nice :)
I mainly use DS as a content generator for Blender. Now in the DAZ importer from Thomas Larsson there's a tool to automatically convert hair to Blender dynamics. I guess what you need for dForce is a similar tool. May be someone with a good knowledge of scripting can write it. The Larsson's project is open source.
http://diffeomorphic.blogspot.it/
http://diffeomorphic.blogspot.it/p/hair.html
I know jack about scripting in DS LOL
However, I'm noticing something: It appears a recent update has made dForce more functional with hair. I'm not seeing it fall apart, even without the pre-treatment. The Riva Updo side curls don't melt like they used to. Set the Dynamics Strength to 0.8 and they bounce. They don't follow the head, though, like when it's tilted fully back and they should swing back. Still, they're not falling apart, so it seems they did address something, probably to fix clothes that melt at the seams. Gotta love positive side-effects :D
Pre-treatment still gives you more functionality, though.
I don't see any update other than the 4.10.0.123 .. the effect you are describing sounds like you turned gravity off. If you can be more specific about the recent update and the tests you did it'll be very interesting. Thank you.
I haven't had the time to actually try it yet, but I look forward to it when I got some time to spare. I ripped it to my desktop so I don't forget about it, I plan on really diving into dforce for hair soon. I like MD better for clothes for now, since I make my own a lot, and I have a workflow I know works.
I'm also getting a Spring calculation in the status window, which didn't happen before. I'm not using the Public Beta (though I probably should since I do have it).
I don't turn Gravity off, so it's not that.
The simplest test is to apply dForce to any hair and run the sim without any adjustments - the standard "let's see what happens and what I need to fix" test. Since none of the strips are anchored into the skullcap, and are in fact separate meshes, they should fall apart, especially the Lustrous Hair, which my videos used. Specifically, the Base section should normally fall completely out. Now it stays put and I didn't do anything to the base figure. All I'm doing differently now is setting Collision Layers to control the layering.
It does look like if given enough time they'd fall apart, even though the ends are not stuck into another mesh, which means they should immediately separate.
OK my first try at following the tutorial, any idea why my hair is acting like there is strong wind from behind? Unless wind is on by default somehow, I didn't turn wind on lol. She is still in her A pose here.
OK, turned collision distance down to .1, that seems to have stopped that windlike issue, now to try some more poses and see how it holds up :)
![image](https://www.daz3d.com/forums/utility/thumbnail/648486/FileUpload/3e/7c86802c3da2417dc7de2fc44e7867.jpg)
Tilted her head back, not looking so great lol. Need to keep tweaking I think.
One more try, looks a bit better. Out of play time now.
If any of the layers (base, top, etc) or the individual "strips" in a layer or between layers intersect they will stick together and entwine as they fight to occupy the same space at the same time. Collision Offset helps with layer separation, but if they start off intersecting as part of their base design, Collision Offset won't separate them. You would have to delete the intersecting polygons.
Also, the outer layer should be a little "thinner" than the inner layers, so that it maintains separation and doesn't try to hold the inner layer in place. You may need to separate the layers using the collision offset by 20s instead of 10s (default is 0.20, so the innermost layer would be 0.20 while the layer on top of that would be 0.40 instead of just 0.30).
Without that exact hair, I can't say for sure what causes it or how to correct it. What hair is it?
It doesn't have a bunch of different layers on top of each other, only in the back. So I have 3 layers, the skull on 1, the backunder on 2 and the rest on 3. Its called Ronya Hair. Since it is so short, I only cut each dynamic part into top and bottom. The last setting used was setting collision offset at .10, the top section dynamics to .85 and the bottom to .98. I also cranked up the stabilization time to 20 and subframes up to 10.
These tests I put the top section dynamic to .9, the at .99, put subframes up 20 14, and collision offset down to .09
Perhaps it's just not a hair that is going to work well with this system, going to give a different one a try today. I really want this to work, as simulating takes far far less time than I spend tearing my own hair out in ZB trying to get hair look like it's affected by gravity and not floating above the skin etc lol. I figured it would be easier and better to get it done on some short hair first, but going to jump to a longer one now, see what happens.
Hmm. It could also bee the construction method of the hair.
"This high quality hair package comes with a detailed 3D mesh made of lots of small hair strands that give this hair real-world volume"
If it's made of anything other than the standard strips, such as tubes, or even skinnier than normal strips, they might not get along with dForce.
In my first test I just applied a gradient falloff and turned off self-collision and it worked quite fine. May be I was lucky with the G8F pack. I didn't try any other hair yet.
https://www.daz3d.com/forums/discussion/223616/help-i-can-t-get-dforce-to-explode#latest
awesome!
Nice. I wouldn't expect short hair to be constructed with too much complexity, though, so they should be a lot easier to work with. Short hair works for a lot of scenes and poses because you don't expect it to move a whole lot, especially reclining scenes with a lot of forward and back motion of the figure. Long hair has always been a problem here because of the limited range of morphs for those types of poses and animations.
Out of curiosity I did a test with the Charm Hair. Always just a gradient falloff so it's a very simple setup. Not bad but not good either. I believe it's because there are too few strands to work with. In the case of long hair this becomes much more visible in the animation.
https://www.daz3d.com/forums/discussion/223616/help-i-can-t-get-dforce-to-explode#latest
Nice. I admit I never tried anything with the weight mapping until just now. I'm kinda disturbed there's nothing on YouTube about them. I was able to figure out how to apply one similar to your example, and may see about combining both methods. I've got an in-progress file where I'm splitting the Charm hair into multiple sections, and individual strips. I'm up to about 75 new surfaces and waiting for it to crash and burn LOL
Still the issue persists that was pointed out in the other thread: strip hair is very limited, and downright ugly once it starts flailing about. No headbanger's ball/Deathklok animations, I guess.