How
can I create good hair physics in my MMD model that I am making from
scratch using Blender and PMXE? Tell me about the Knife tool in PMXE.
Making Your First Model Part 19 Adding Hair Physics
Before we get started . . .
Let’s troubleshoot this ankle.
Bone
panel settings for all IK Bones
MigitsumasenIK and HidaritsumasenIK are identical, down
to limiting the angle the way Bandages told some person to do in dA
comments that I was lucky enough to stumble upon.
Remember, the left leg
works perfectly. The right leg doesn’t.
But even with pasting all of the information panels side by
side, I can’t find the difference!
By both verifying that the Target and the IK are in the
same place via the bone panel and making
sure her feet are in their “default pose” in transform view . . . I got
rid of the possibility that the location is the problem.
I think I’m going to be a big dummy and delete her right
IK bones and redo them via mirroring to copy the correct settings.
As a note, if you wish to mirror bones . . . Make sure
to not include any invisible bones. Sometimes it will work . .
Sometimes it won’t. And I don’t know why.
Here’s another thing that I don’t
know why it is . . . This time, her target bones
were mirrored as well! Let’s weight the bones we just added.
Both
ankles working properly!
I did it!!! I DID IT!!
But here’s something oddly demoralizing about PMX. At
first glance, there was literally nothing wrong.
Actually, at close inspection, there was literally
nothing wrong. So, what did mirroring the bones do
. . . ? I don’t know, but it’s fixed! So let’s move
on.
I came here for physics, not
IK
Yes! Yes . . . Of course . . . Let’s get on to physics.
In the last installment, I showed you how to add bodies.
I hope you didn’t “nope out” before that point, because we’re just
going to assume you have them right now!
So, in my frantic search to get the skirt plugin to work
. . . I found out about the “naifu” tool. (The more frustrated I get with Japanese
freeware, the more Japanese words I use because nothing quite says
“frustration” like tapping into a language you wasted five years on.)
So, if you have the English version like me, it’s up
near all of the other buttons that make the weight and masking box pop
up, and it is labeled “Adv”. Click that, and it will pop up.
Now, if the tutorial that showed me how this worked
wasn’t lying . . . Choosing the series setting in the bone tab will
allow me to plop down all of the bones I need without having to
actually having to parent / child them. Before we get started, let’s go
down to the bottom of the view window and pick the green thing next to
the blue for bones circle. . . . It goes away after you start the
plopping down of the bones, but it’s nice to have a basic idea.
–
We want to make sure that each bone is weighted to at
least two edgeloops. I have no proof for why that would possibly be
better, but it just sounds better.
How would physics help ya if each bone only had a tiny bit
of hair to control?! Let’s mash start on the advanced editing window
and start plopping down bones.
It almost seems as if this software is
purposefully obtuse! Just select that chain and
shimmy it into the right place.
Before we add physics, let’s weight the braid. Then
let’s change the name from the default new bone name to “braidx” with
the “x” being the correct number. Make sure to parent the first bone to
the head. Now, let’s select all of those bones and go to “Edit >
Bone > Create Body / Joint”.
Now, go to the “Joint” tab and make sure to fix the
first joint. As of now, it’s connected to nothing. Make sure to connect
it to “頭”.
Joint
pane with first joint linked to the head body
Now, let’s jump over to the “Body” tab and select all of
the braid bodies. We want to set this to any group that isn’t already
used. We’ll just use two.
Body
tab with braid bodies assigned to group 2
Now . . . We can test. Pull up the Transform View.
To turn on real time physics, mash the yellow triangle
up in the transform view toolbar.
If you followed my instructions . . . you should have a
twitchy mess. If you didn’t, there’s a chance her hair will be on the
floor. Remember to link the first joint with the head!
Now, let’s try and get her to stop twitching. Let’s go
into the joint tab and mash in random numbers. Let’s make the
rotate nearly non-existent and the move tab pretty . . . existent.
Let’s ask a philosophical question.
If every box you are able to put data into is null,
should the braid twitch as if it is experiencing a seizure?
No. But it does.
Then realize you were a big dummy and that the body
tab also has control over the
braid.
First, set the group the braid is in to a non-collision
group . . . Because I’m a big dummy that didn’t think you needed to do
it, ’cause, ya’know, hair doesn’t pass through itself!
Body
panel with non-collision group set to two
In joint, I set Spring Move to .5 , 1 , .5 just because
it seemed like a good idea.
You know what else seems like a good idea?
Fringe physics. Those tendrils on the side of her face?
Let’s make them move with her!
Place the bones, weight them, then add the automatic
bone / joint . . .
Now, there’s a chance that after doing everything
correctly . . . It’s a twitchy mess. That could be because your group 1
bodies are too close to your group 2 bodies. In this case, it would be
the head in the way. . . or it might be the last body in the fringe.
Adjust the size of the bodies until it looks better.
And then mash in random values until it looks good.
Hopefully, you’re happy with your result! If you aren’t . . . mash in
random values.
Now . . . For her bangs . . .
Let’s add a nose body so that it won’t collide with her
nose.
After adding a nose collider, I tried to use this
tutorial for her bangs.
After struggling quite a
bit, I finally got to a point where I was ready to
test it out.
Well . . . It’s not perfect . . . But it
is close enough!
Next time, we’re either going to tackle either the skirt
physics, facials, or textures.
I’m not sure yet . . . I sent out an email blast to all
LearnMMD contributors to see if anyone could help me with the skirt!
So tune in next time when we . . . do . . . things?