Azusa


This is my first fully original character model. She’s not done, but she’s getting to the point where I’m regularly happy with my renders. Not even sick of working on her yet, so she might not be done for a while.

Going for a glossy comic book look, and where I have that, I’m happy; where I don’t, I guess I’ll keep working at it.

Today, rather than fix her weights (look at that neck), remake her volume-losing facial expressions, learn rigid body physics to get her hair to do something, or figure out how to prevent the tangent-related artifacts from 166-fold scaled normal maps, I made some almost imperceptible changes to her hair textures and made some glasses.

I came to Blender from Miku Miku Dance and I’d love to be able to get her working in MMD as a going-away gift to them, but that’s going to mean a lot of bone work trying to translate her constraints into that animation system, so I don’t know when I’ll get around to that-- especially considering she probably won’t done until I’m utterly sick of working on her. But if, after all that, I’m still not sick, I’m hoping she can be the model that takes me into Unity.

I’ve been doing a lot of rigging and figured I could write about my thoughts.

I started learning 3D modelling and animation via Miku Miku Dance, a ten year old Japanese program. In the West, the community is dominated by teenagers. Any serious documentation is in Japanese, which I neither read nor speak. Rigging tools are limited. The local axes of each bone is set to the global axes in the default pose. There are IK bones and copy rotation/copy translation tools. That’s it.

At first, like everyone probably, I thought about bones as being like bones. That’s the paradigm. It’s a long thing that moves your model. Put them where bones are.

The first time that changed was when I realized that the end points of bones don’t matter. Make a full FK rig without any constraints, move all the tails to some arbitrary position, and animate in global orientation, and the tails are meaningless. So then I stopped thinking about bones as bones and started thinking about them as joints. It’s not the forearm bone, it’s the elbow bone. It’s the center of rotation that matters, not the tail.

When I started rigging in Blender, for a while, I went back to thinking about them as bones, because Blender makes you give them tails, and those tails are actually useful! But that was dumb of me. They’re still not bones. I started to discover this when examining the Pitchipoy facial rig in Rigify.

Bones aren’t anything. They’re groups of deforming vertices. In Blender, there’s an extra little pocket where you can store information, the tail, which is just four little floats. And just like when I was writing shaders and I started to realize I can use textures to store any kind of data I want, I realize there’s no difference between addUV and vertex color, it’s just how you use it-- just like that, I realized that you use that tail vector to store any kind of information you want. It doesn’t matter where the tail points, what matters is what information that tail vector contains, and how you use that information.

But it’s hard to get rid of metaphors all the way. So I still think of bones as something. But I think of them as lumps of flesh.

People will tell you to study anatomy, watch how the femur moves in the socket. But when we watch an animation, we don’t see the femur moving in the socket. We see the flesh that is pulled by that femur, constrained by adjoining flesh. That flesh isn’t anchored to any socket. It maintains its volume. It slides around. It doesn’t rotate around a fixed point like a bone does.


How do oblique muscles move? What point do they rotate? They don’t rotate a point. They connect points.

I no longer choose deforming bones for their centers of rotation. Will that selected oblique deformer move in any useful way according to its chosen axis? No. I chose the axis to tell automatic weights the general shape of the lump of flesh I want to animate. That’s how I place deforming bones now.

When I place deformers, I test them out. Scaling is good. Just to see their margins. Then I put a cursor down and start examining different centers of rotation. When I get a good center of rotation, I place a non-deforming bone there-- tail doesn’t matter yet-- and I parent the deformer to it. This means that my model changes shape-- its obliques don’t just rotate, they expand out of the body. Some people tell you not to do that, because they’re stuck thinking about bones as bones. Bones don’t expand out of the body but lumps of flesh do.


This is the cross-section of the groin->hip from a very high quality model, by a modeller I have supreme respect for, a modeller that has been modelling for decades. It’s not newbie shit. But compare the cross section of the groin in default pose with the cross section of the groin with the leg rotated 90 degrees out in Blender’s Y axis. Legs don’t do that. Cartoon legs don’t even do that. The cross section stays mostly the same in real life. But modellers have mostly thrown up their hands with dealing with this problem.

I read all these people that tell me that rigging is easy, it’s just tedious to do the fingers. Then where is the model that solved this problem? It’s not unsolvable. Then those same people get defensive. They tell me that people can’t do that anyways. Which is not true (just google “splits”) and irrelevant. Because we’re not making real people. Most of the time, we want models that can do MORE than people can. We’re playing with the fantastic.

Like I said, this isn’t an unsolvable problem. But it’s not solvable as long as you’re stuck thinking that bones are bones, when they’re not.


We all know this problem. People try to solve it with weight painting, which doesn’t work very well, they try to solve it with topology, which seems to work well so long as you limit yourself to 1k vert models, they try to solve it with shapekeys, and I don’t know how well that works, I haven’t tried it. It’s solvable with rigging. I’m not doing it perfectly yet though.

What is the distance between the center about which your elbow rotates, and the inside of your elbow? It is not a fixed number. It depends on the angle. If it doesn’t depend on the angle, then you get the cross-section that squishes, and where did all your flesh go? Did it all squish up into your bicep?

So why not make an inside of the elbow bone? And rotate it about a different center, so that as it rotates, it also moves. Place the bone for good response from automatic weights, rather than for some imagined anatomical reason, and you don’t even have to do any weight painting.

Except you have to animate it. But you don’t, if you store information in the tail of its non-deforming parent. How should the parent point? Figure that out, and then place the tail at that point in the default pose. Its tail is meaningless except to store that extra data that can be used to automate animation.

Interesting take on rigging and I agree on most points.

I believe that weight painting and topology will only take you so far, then you have to come up with some rigging “tricks” to make it all work. Sometimes I’m successful, and other times the rigging will throw me for a loop and I have to re-think things, which is why I’m leaving the hard stuff (rigging humans) to the experts.

I’ve seen some Blenrig-rigged models that are really well done, but they use deform cages – I’m not going there any time soon. I’ve been making robotic-type characters lately and it’s a lot easier for me to rig those since they have limbs that are rigid and rotate around a fixed point. Even then, I sometimes have to work hard to get things to pose correctly. Rigging is difficult, yes, but I really love how Blender is very flexible in this area and once you get the basics down it makes things a lot easier to manage.

The rigging is often still throwing me for a loop :slight_smile: I’m not talented, but I’m hard working and ambitious, and I know I’ll get there eventually.

I’ve seen some Blenrig-rigged models that are really well done, but they use deform cages – I’m not going there any time soon.

Yeah, it’s been the deform cages (and the need for the add-on) that has made me reluctant to use Blenrig as well. But I ought to give it another shot. More and more, it looks like I’m using Blender models for Blender rendering, so I might as well take advantage of all possible tools. There’s just a part of me that keeps whispering that when I want to use Unity, it will be easier to code Blender constraints than it will be to code a mesh deform.

My experience playing with Rigify, as well, leaves me wary of anything that tries to do it for you. I get scared that rigging add-ons are going to make a lot of assumptions about what I’m going to do. I imagine I could spend a long time learning Rigify and get it to do what I wanted, but at that point, I might as well spend a long time just studying basic Blender and getting it to do what I want instead.

I’ve been making robotic-type characters lately and it’s a lot easier for me to rig those since they have limbs that are rigid and rotate around a fixed point. Even then, I sometimes have to work hard to get things to pose correctly.

It’s definitely nice not to have to worry about smooth weight transitions, but a lot of mechanical rigging can be very hard. It’s not just about getting motion that matches the mesh, it’s about getting controls that are simple and still powerful enough to get your model to move in the right way. I think rigging is always going to be as hard as you let it be. If you have low standards, it will be easy. If you struggle to make everything as good as possible, it will be hard.

I really love how Blender is very flexible in this area and once you get the basics down it makes things a lot easier to manage.

Oh my god yes. I’m in heaven. I did what I could with MMD, using IK bones to emulate damped tracks, using IK bones located off their targets to initialize bones onto custom local axes, cam-shaft structures to turn rotation into translation, all sorts of stuff, but there’s a complexity ceiling, a point where I lose track of what I’m doing and say, “That’s it, it’s not worth trying to do more because it’s just too hard to integrate it all, I’ll just end up breaking everything else.” With Blender’s tools, that ceiling is so much higher. I haven’t yet had an idea that’s too painful to test.

I’m building my rig for myself, mostly. I’m not an expert animator. I hate FK because I think in terms of foot position, not knee position, and because the foot positions is dependent on the knee transform. But IK controls don’t leave me enough control of the knee. Pole systems are okay, but don’t work great when you get IK chains > 2 bones, and there’s something unintuitive about them. I think a good example of an animation where both of these systems fail would be a character dropping to their knees and toes, then shifting their weight between those four points. Trying to keep the four points on the floor with IK+poles or with FK is very difficult.

How I want to animate is to place the hand, then place the elbow and clavicle if I need to. Only if I need to. But to do this, I need a good first approximation for the elbow. So that when I place the hand, I can actually see what it’s going to look like. And to prevent bad interpolation through weird angle positions, or screwed up models when I forgot to set the elbow.

What I’m doing is using an IK chain as a guideline, and I decided to make an automatic pole for that IK chain. I’m not totally sure it was necessary :slight_smile:


So to start out with my bones, there’s not a lot I know for certain. For these legs, I know where the pelvis is, and I know where each foot control is. From that, I try to figure out a smart place to put the pole. I do that by drawing a triangle: from the pelvis to the ankle (foot control base) to the toe (foot control tail). By placing a bone on side of the triangle and aiming it through the other side of the triangle, I can place a pole marker for my IK chain to get a good idea of which way it ought to point. Then, if the animator wants, they can change the knee position so it rotates differently.

What I found with this was that it worked pretty well, except for when I was rotating in Blender’s Y axis, which I’m going to call foot roll, even though that’s confusing. The triangle was changing shape when the foot rolled. In real life, that kind of roll is very limited, but it’s best to have reasonable output through even impossible angles for purposes of interpolation-- people don’t really care if a model does something impossible, as long as it’s just a little impossible and as long as the model doesn’t stay that way.

So instead of drawing the triangle all the way to the toe, I drew it to a bone parented to the toe, but above it, so that when the model rolls its foot, the triangle angles outward or inward, and the IK guideline adapts.


I like these kinds of things where the precise placement of a bone gives a model personality. I could make another copy of this model and just move these markers, and the model would move slightly differently. I think the holy grail of rigging is a model that you can feed canned animation-- made for a completely different model-- and it not only does it right, it adds its own personality to it.

The elbow/knee pole issue is perhaps my biggest hurdle when rigging – I just can’t seem to get them to do exactly what I want. I try parenting them to different bones to see if I can get the knee/elbow to point like I want, but that never works for me. I guess I need to study these things more, but with a full-time job it’s tough for me to find the time.

I don’t think the “default” structure is bad. It’s fine to say to the animator, look, you want control of the knee? Then place the pole. If you wanted, you could stick it on a transform-locked rotation-only parent, but there’s always going to be something the animator has to use if they want control.

Where I have noticed that it doesn’t work well is with >2 bone chains. If there’s a good way to get good control of a clavicle->humerus->forearm->hand with an IK target + pole, I’m not aware of it. Wouldn’t mind a link though, if it’s done well somewhere, it’s easy for me to overthink things and make them harder than they need to be.

My own clavicle is one of the last few things bugging me. It’s right automatically, most of the time, but I can’t figure out how to get the roll just right. It’s like it should point a face at the elbow to a certain extent, but depending on whether the arm is raised or lowered. Might be able to do it with some careful use of constrained marker bones, I don’t know. But I guess it’s not the worst thing in the world if you have to pose the clavicle sometimes. Just nagging at me because it seems like it ought to be possible. I might work on something else for a while to give the problem some brew time.

Rendering a physics test for the earrings. Has been more of a pain in the ass than I thought it would be to try to copy transforms from rigid bodies. Can’t figure out a way to get the rigids to follow their ultimate bone parents (via constrained rigid “parents”) in pose mode (no animation) without screwing up the physics, which is disappointing, harder to pose when your ears won’t look right until you hit play.

OMFG drivers.

I should’ve learned those a long time ago. Of course, a long time ago, I probably wouldn’t have the experience to be able to learn them as quickly as I did today… Takes time for a lot of the knowledge to sink in for me, to sink in enough that I’m ready for the next thing. I’m still struggling with rigid body physics, but I’ll have that soon; after that, I think I need to learn compositing.

Made a little bone switch, used it to switch rigs. Unfortunately, can’t figure out how to get video on here, it doesn’t even like my gif file. Time to give up on that I guess.

Anyhow. All I’m doing is driving the influence of some copy transforms constraints with my switch. So I’ve basically got a standardized skeleton driving my control bones. The rig’s still nice, because it does things smart, so I don’t have to tweak things like twist bones or poles, don’t have to copy transforms to a hundred deform bones, just to a few control bones, and the main rig fills it all in for me. I’m currently using that to look at some canned animation designed for the standardized skeleton.

But one of the reasons I’m really in love with this is because it means I don’t have to make a one-size-fits-all rig. If I need a rig that behaves differently-- say, a rig for lying down-- I now know that I can do that relatively easily, well compartmentalized, and interpolate smoothly between rigs, without any scripts. That’s a nice feeling, and it frees me from struggling further with a rig that’s nearing my personal complexity ceiling.

Hi, :slight_smile: interesting thread.

Videos have to be on an external channel, like youtube or vimeo and here you put in the link only.

Happy publishing :slight_smile:

@bandages – you are so much more knowledgeable than me in regards to rigging, but I find it fascinating the ideas you are coming up with. Keep it up.

Oh, thanks. Not worth clogging up Youtube with WIPs though.

Thank you, that’s very nice to hear. I get discouraged sometimes and it’s kind words like yours that keep my head above the water.

Added physics to hair today. Just rigid body, capsule colliders. But it brought my model to life. I’ve been staring at the same hair mesh for months, gradually getting sicker and sicker of it, and suddenly her bangs are sweeping across her forehead, bobbing when she nods, and I’m re-energized, I’m in love again. It seems to me like there are two things that are important for character models, hair and eyes, and if you get those right, you can afford to get everything else wrong. Mine aren’t right yet but I’m getting better :slight_smile:

There isn’t a lot more that I need to do for release. There’s always more that I can do, of course; if I had the energy, I could work on a single character for a decade. But I need something fresh (well, fresh-ish, I’ll use what I can from this) to experiment with some new ideas. Mostly just have to rig some clothes. Maybe that’ll be more work than I’m anticipating.

to easily share videos just for a forum you can use dropbox/google drive/OneDrive or whatever they come with tons of storage to plop in some WIP videos

Not a bad idea, thanks!

In other news…

I think I’m done.


I’ll sit on this for a day before uploading it to Blendswap under a CC-SA. A day’s just to let myself catch any stupid last minute errors. If anybody wants to take a look and let me know about anything dumb, there’s a copy at http://www.mediafire.com/file/a5ga1wy9ojbgot6/0404Azusa.blend .

Edit: hmm, maybe this site doesn’t handle transparent pngs very well. Looks fine elsewhere, I’m not worried.