Monday, August 26, 2013


Yes please watch this now. It is sweet.

The Last of Us. Rigging Video

I am watching the video now and right off the bat.

They talk about modeling the eyes half closed and the mouth open... I am glad this is finally taking off.

Judd Simantov: Rigging, The last of us. Youtube Link.

Also they talk about using joints more like muscles. I feel like CG artists are finally pulling themselves away from how things are traditionally done and looking for more realistic ways of rigging.
Every time I watch a rigging video the joints are always put center of mass. I am sorry but your spine and neck do not run down the middle of your body. they run along the back.

The wrist setup was interesting. I have been fighting with wrists my entire rigging life and this has given me an idea. I will keep everyone posted. All in All good video and I understand why they never go into the actual setup, money, but it's enough to get my brain rolling on a couple new setups.

Wednesday, August 14, 2013

I shall call him, "Fizzle"

New project I'm working on.


Look at him, love him, fear him.

The little things that kill.

Just a couple of interesting things I found out while rigging batman.

  1. If you like a perfectly clean rig NEVER turn on show auxiliary nodes. You will find yourself staring down a couple hundred un-named nodes, like group IDs and trying to discern where they come from and how to name them will drive you up a wall. In my opinion just leave the aux nodes alone unless you need to edit them.
  2. DO NOT Rigid bind unless you are sure you need the functionality of a rigid bind. A parent constraint works just as well and with less nodes than a rigid bind. For a parent it creates one node, for a rigid bind it creates at least 3 if not more depending on how you apply the rigid bind
  3. Dual Quaternion Skinning weights are not always your friend, specifically on the leg to groin area. The problem is the skin from the groin to leg stretches linearly not rotationally so you get unwanted bulging on the inner thigh when the legs spread apart.
  4. Never Ever Ever assume that the way you were shown was the right and only way to do things. If we never question we will keep repeating the same unseen mistakes.
Just a couple things. Bon Appetit.

Tuesday, August 13, 2013

Retopologizing the world one quad at a time.

As I flipped through the 3D magazines at Barnes & Noble years ago and saw all the incredible sculpts being done on zBrush I was amazed by the level of detail and always wondered how it would look animated and why Pixar wasn't using this high frequency detail in all their work. I decided to look into it and as I pulled back the layers I found that most, if not all of the sculpts were completely un animatable, and I was kinda bummed when I learned the reasons.

High resolution sculpted mesh until very recently had major problems it was fine to sculpt but the topology was usually just boxy with no edge flow defining shape. This problem is magnified by the fact that you needed to use the lowest possible level to animate because less geometry in a mesh the easier to setup, rig, and animate. This problem is that when you reduce a non properly apologized mesh you almost immediately lose the ability to deform the mesh while still keeping the shapes and mass. Soon enough I was looking at these amazing sculptures, yes they are just sculptures if they cannot move, in these pieces of software and saying to myself.

"This is great but it will do nothing more than just sit there lifeless in a render and look pretty."


Thats right folks! Now when you want to use your high resolution sculpted mesh for animation purposes look no further than,
Thats right now mudbox and maya come built in with re top tools, I'm starting a new catch phrase, thats right re top tools.

This is actually the answer to the dreams of a lot of people including me. Now I know that zBrush has been working on this for a while but ignore that for now.

The ability to retopologize is an amazing step in allowing the art to take forefront and the topology be worried about after you finish. A great example: You have multiple sculpts for your boss and each uses a different set of scull spikes or ears that would if modeled involve 3 different sets of topology. Instead of re modeling then re sculpting you can sculpt layer each change from a base mesh then re top the mesh after your boss has chosen the final look. He chooses the one he likes then you allow the re top tools to do the rest, kinda.

So to all of you sculptors out there.

Do not make the riggers cry as we cannot rig pre posed models easily.
You don't want to make the riggers cry... bad things happen, like your model gets his head shoved up his sphincter.

Catcha on the flipside of the normal.

Monday, August 12, 2013

Gimme` a "T"(pose) Gimme` an "A"(pose). The battle of the Poses.

Ok this has been bugging me for a long time now and it falls into the category of , why are we doing it this way.

The T pose and the A pose.

T Pose : This pose is the standard pose. Arms straight out 90 degreees from the body, hands are palm down. Legs are straight and feet point straight ahead.

Pros :
Easy to setup joints and other rig pieces that function or setup easier or better when limbs are aligned with the world axis. Easier to model into T pose as limbs are once again straight.

Cons :
I will go area by area.

  1. Forearm: The forearm is modeled twisted and can cause issues when un twisted during rigging and animation.
  2. Shoulder and Clavicle : When modeled in T pose the clavicle is slightly raised and the deltoid is engaged and squashed. This means that you must model the deltoid knowing that it is at its maximum squash as the shoulder only rotates to 90 degrees before the clavicle is engaged. This one is a double edged sword because on larger characters with fat and skin, un squashing the underarm a little can help when modeled in the T pose and puts the arms higher.
  3. Groin : How comfortable are you sitting with your legs completely straight? I know I'm not because it squishes my groiny bits together. T pose squashes your inner legs and groin together and also in this pose the skin begins to self collide and must be modeled as such. In modeling collided flesh you flatten out the area that must later be given volume when the legs are spread.

Now onto.

A Pose : Limbs are modeled in a more relaxed position minimising any squash or stretch to any part of the body. Arms are at a 40 degree angle and slightly forward. Elbows slightly bent.

Forearms: This is a biggie as there are 2 ways to model this. As the arm twists just less then 132 degrees we will call the middle point 67.5 and zero is with palms forward and forearm bones completely un twisted

  1. Anatomically un twisted: Modeled palm forward this puts the forearm in its un twisted form the Zero degree twist state. The bones are straight and the muscles are straight but the arm geometry will need to twist 132 degrees for the maximum human rotation, thus twisting the geometry more than the next option.
  2. Anatomically neutral : This is when you model the arm at the 67 degree mark and slightly twist the forearm but model the geometry straight so that it deforms minimally from median to max/min as it is only going 67 degrees in either direction.

Pros : 
When rigging and deforming the mesh starts from a more neutral state. The mesh must go through less deformation to assume opposite maximum poses e.g. : Arm from straight out in T pose to straight down along the body.
Also this means that the joints need to rotate less to assume certain positions. This means less flipping or other bad joint behavior.

Misc Pros :

Cons : 
Personally the hardest thing is not being able to model straight out, having things in an A pose means making sure things flow at the proper angle the whole time and not letting things get curvy.

Other differences :
Eyes: The eyelids are modeled in a closed position, this is the natural state of eyelids as when open they fold onto themselves and into the brow or cheek. This makes rigging easier as the eyelids line up  perfectly during a blink and not in the hands of the rigger to make the blink work by deforming the eyelid to eyeball line.

Mouth :  Also modeled in the slightly open position so that the upper and lower lip do not touch. This allows for easier rigging of lip collisions and squashing as when the jaw is completely shut the lips push together slightly which is not a neutral state.

All of this is just my opinion because at the end of the day it is art.

Here is a grain of salt.

Wednesday, July 31, 2013

Batrman Beyond Google drive file.

Still waiting for Creative Crash to upload my file. In the meantime!