Stumbling Toward 'Awesomeness'

A Technical Art Blog

Monday, March 5, 2018

uExport: A Simple Maya Tool for Exporting to UE4

I posted the skeleton of this a while ago in the post “Why does everyone write their own FBX exporter?“, many people asked me for the exporter over the years and I sent it to them. Since that post, uExport has become the way we get all characters into UE4 at Epic. It’s been crazy to see this little fledgling tool, developed largely in spare time to allow us to export non-A.R.T.v1 rigs (like the kite and deer in the first demo I worked on at Epic) become something we rely on so much.

It’s here on GitHub, there’s a lot more info in the readme.

posted by Chris at 9:05 PM  

Monday, February 5, 2018

Python Style Guidelines: Consistency is King

I wanted to touch on something that comes up every once in a while: The PEP8 Style Guide. When you’re first getting into Python, you are messy as hell, and as you gain more experience, you start understanding the need to adhere to basic standards for code style. However, just at that moment, I see people completely miss the point of PEP8, and go down this path of assuming that anyone not adhering to PEP8 is doing it wrong.

That’s not the case, and without really understanding PEP8, you could easily fall into this trap.

PEP8 Is About One Thing

PEP8, as defined by itself is about one thing over all CONSISTENCY. Literally the FIRST THING PEP8 says after the Introduction this:

But people don’t really want to pay attention to that, I mean you’re looking at the PEP8 docs to be told how to write code, not that you might be in a situation where you should ignore their standards. But the most important message they impart is to be CONSISTENT.

Python Application APIs

In the VFX and games industry, most of the Python we write is inside existing applications through exposed C++ APIs that have been wrapped in Python. Often people wonder why these professional applications don’t adhere to the PEP8 Style Guide. The answer is simple, they are wrapping existing C++ functions, for CONSISTENCY they use the same camelCase function name as in C++, not lowercase_underscore following PEP8. But why?

  • It allows someone who knows the C++ function to immediately know the Python function
  • It allows a user who is accessing a new method added to an existing class to not have to wonder about it’s naming
  • In documentation, the often human-written explanations of functionality are often completely usable for Python docs

At Crytek, when we implemented Python exposure in the engine, we had to think about this and I noticed that almost all applications with C++ Python APIs were consistent with the C++ style/naming. Here are some examples:

  • Maya: cmds.setAttr()
  • 3dsMax: GetLengthSquared()
  • CryEngine: Alembic.playSequence()
  • PySide: QtGui.QMainWindow.setCentralWidget()
  • Modo: scene.ItemLookup()
  • Houdini: point.setAttribValue()
  • MotionBuilder: actor.setDefinitionScaleVector()
  • Substance: sbsexporter.getExportedDependencies()

Are all these developers stupid for ignoring the ‘Jesus book’ of Python Coding Standards? No, they are actually following it! Remember: CONSISTENCY!

Interpretation of The Good Book

Let’s take a look at some times when you should or shouldn’t follow the good book. These are just my take on things, I would love to hear your take in the comments. Also, this post mainly centers on naming standards, it’s usually best to decide as a team which PEP8 rules you want to follow and which don’t make sense.

I am generating a python API that wraps my C++ code

Be consistent with your naming standards that are already set. Even Python itself has multiple libs that don’t follow PEP naming just to be consistent with existing work. This one was covered above, so I’ll move on.

I am writing code that fits within an existing Python API

Be consistent with that API. If you’re in Maya or using PySide, you’re forced to use their function names. Please don’t have half your code with their function names and then half the code with your own PEP8 function names, this is really inconsistent. Not to mention when you override or reimpelent existing functions,  you’ll find yourself typing something like “on_close”. Another annoyance will be style enforcement in IDEs, if you use PEP8 you are probably enabling style enforcement, and you’ll a ton of false warnings from the API you’re using.

I am writing a standalone Python application or package

Use the PEP8 Styleguide if it makes sense. But don’t feel the need to refactor all your code, again, just take a look at Python’s own threading or logging modules, they don’t adhere to their own standards because it didn’t make sense.

I was pointed to the blog of Eric Husler, a veteran VFX and game pipeline programmer who covered this very topic in his post PyQT Coding Style Guidelines, his sentiments also follow the above:

“While you may be programming in Python, you’re creating Qt classes – inheriting their conventions along the way. When adding a new method, you should keep the consistency of Qt – that way someone working with your widgets doesn’t have to think about whether a method came from Python, and so use underscores, or came from Qt itself, and so use camel humps.”

He offered good examples of what we are discussing above, here is a widget with custom functions consistent with QT:

widget = MyWidget()

Here is a widget that has custom functions that adhere to PEP8 style but are inconsistent with QT:

widget = MyWidget()



posted by Chris at 9:39 AM  

Monday, November 13, 2017

The Mighty Message Attribute

I recently had a discussion about storing relationships in Maya, and hadn’t realized the role of the message attribute wasn’t this universally cherished thing. In previous posts entitled ‘Don’t use string paths‘, or ‘Why Storing String Refs is Dangerous and Irresponsible‘ I outlined why this is the devil’s work, but in those posts I talked about the API, PyMel and Message Attrs. I didn’t really focus on why message attrs were so important: they serialize node relationships.

For quite some time I have advocated storing relationships with message attrs. At the Maya SIGGRAPH User Event, when they asked me to speak about our modular rigging system, I kind of detailed how we leveraged those at Crytek in CryPed.


I am not quite sure when I started using message attrs to convey relationships, I’m no brainiac, it could have been after seeing this 2003 post from Jason Schleifer on CGTalk:


Or maybe I read it in the Maya docs (unlikely):

“Message attributes only exist to formally declare relationships between nodes. By connecting two nodes via message attributes, a relationship between those nodes is expressed.”

So why does Maya use this, and why should I?

As you read in the docs above, when Maya wants to declare a relationship between a camera and image plane, they do so with a message attribute that connects them. This is important because this bond won’t be broken if the plane or it’s parent is renamed. As soon as you store the string path to a node in the DAG, that data is already stale.  It’s no longer valid.  When you query a message attribute, Maya returns the item, it’s DAG path will be valid, regardless of hierarchy or name changes.

Jason’s example above is maybe the most simple, in my image (a decade later) you can see the messages declaring many relationships abstracting the character at three main levels of interface, Character, ChatacterPart and RigPart. I talked about the basic ideas here in a 2013 post about object oriented python in Maya.

Though Rob vigorously disagreed in the comments there, I am still doing this today.  Here’s an example from the facial code we released in EPIC’s ARTv1 rigging tools some time ago. The face is abstracted on two levels, the ‘face’ and the ‘mask’, here I am only displaying the message connecting them:


By using properties as described in that previous blog post, below I am accessing the system, creating a face instance, walking down the message connection to the mask node, and then asking it for the attach locations. It’s giving me these transforms, by querying the DAG, live:


So, that property looks like this:

    def attachLocations(self):
        return cmds.listConnections(self.node + '.attachLocations')
    def attachLocations(self, locs):
        for loc in locs:
            utils.msgConnect(self.node + '.attachLocations', loc + '.maskNode')

Setting the attach locations through python would look like this, and it would rebuild the message attrs:

face.mask.attachLoactions = ['myLoc1', 'myLoc2']

Working like this, you have to think hard about what a rigger would want to access at what level and expose what’s needed. But in the end, as you see, through python, you have access to everything you need, and none of the data is stale.

How and when to use strings

There are times when the only way you can store a relationship is by using a string in some fashion. Here are some situations and how I have handled them in the past, feel free to leave a comment with your experiences.

  • Maya can’t store a relationship to something that doesn’t exist (has been deleted). It can’t store a relationship when it’s not open. In these situations, instead of storing the name in an attr, I stamp the two nodes with a string attr to store the relationship, then you query the world for another node with a matching stamped attr.
  • Many times you need to feed your class an initial interface node to build/wrap. Instead of feeding it a string name, you can query the world for node type, in the Ryse example above, the rigging and animation tools could query’CryCharacter’), this would return all characters in the scene. This means all rigging and animation tools needed a common ‘working character’ combobox at the top to define the character the tool is operating on. If you don’t have a node type, you can use a special string attr to query for.
  • Sometimes you’re like saving joint names to serialize skinning data or something. You can use message attrs to play it safe here as well. Some pseudocode: For character in characters, if character identifier matches file on disk, for mesh in character.meshes if mesh in file skin it. For joint in character.joints if in file, add them to the skincluster, etc. Here you’re validating all your serialized string data against your class which is traversing the DAG live.
  • Message attrs can get SLOW if you’re tracking thousands of items, you should only be tracking important things you would want later. In CryPed, when we wanted to track all nodes that were created when a module was built, we would stamp them all with a string attr that was the function name that built the module. To track this kind of data HarryZ at Crytek had the pragmatic idea of just doing a global ls of the world when a buildout started and then one at the end and boolean them out, this caught all the intermediate and utility nodes and everything generated by the rigging code.
posted by Chris at 6:10 AM  

Wednesday, July 5, 2017

Skin Weights Savior

Lots of people were interested in Deformer Weights and ways to save/load skin weights faster. Trowbridge pointed out that the API now allows for setting in one go, no loops needed, like the C++ call. Aaron Carlisle on our team here at Epic had noticed the same thing. Aaron took the time to write up a post going over it here:

Using GetWeights and SetWeights in the Maya Python API

Also, it looks like you can get/set blind data in one go now… 😮

posted by Chris at 11:20 PM  

Tuesday, October 25, 2016

Skin Save/Load Tool


Some had asked me to package the previous code together into a tool. I was reluctant because there were many situations where it just didn’t work. I made something to aid a discussion on the beta forum.

However, accidentally, I seem to have the code working. It’s odd, but the solution to skinClusters not being paintable after deformerWeights was used to load them was to try and make another skinCluster (which errors, but fixes the Artisan issue).

I have tossed this on GitHub here:


posted by Chris at 10:46 AM  

Monday, September 26, 2016

Save/Load SkinWeights 125x Faster

DISCLAIMER/WARNING: Trying to implement this in production I have found other items on top of the massive list that do not work. The ignore names flag ‘ig’ causes a hard crash on file load, the weight precision flag ‘wp’ isn’t implemented though it’s documented, the weight tolerance flag ‘wt’ causes files to hang indefinitely on load. When it does load weights properly, it often does so in a way that crashes the Paint Skin Weights tool. I have reported this in the Maya Beta forums.

Previously I discussed the promise in Maya command ‘deformerWeights‘. The tool that ships with Maya was not very useful, but the code it called was 125 times faster than python if you used it correctly..

Let’s make a python class that can save and load skin weights. You hand it a few hundred skinned meshes (avg Paragon character) and it saves the weights and then you delete history on the meshes, and it loads the weights back on.  What I just described is the process riggers go through when updating a rig or a mesh _every day_.

Below we begin the class, we import a python module to parse XML, and we say “if the user passed in a path, let’s parse it.”

#we import an xml parser that ships with python
import xml.etree.ElementTree
#this will be our class, which can take the path to a file on disk
class SkinDeformerWeights(object):
    def __init__(self, path=None):
        self.path = path
        if self.path:

Next, let’s make this parseFile function. Why is parsing the file important? In the last post we found out that there’s a bug that doesn’t appropriately apply saved weights unless you have a skinCluster with the *exact* same joints as were exported. We’re going to read the file and make a skinCluster that works.

#the function takes a path to the file we want to parse
def parseFile(self, path):
    root = xml.etree.ElementTree.parse(path).getroot()
    #set the header info
    for atype in root.findall('headerInfo'):
        self.fileName = atype.get('fileName')
    for atype in root.findall('weights'):
        jnt = atype.get('source')
        shape = atype.get('shape')
        clusterName = atype.get('deformer')

Now we’re getting some data here, we know that the format can save deformers for multiple shapes, let’s make a shape class and store these.

class SkinnedShape(object):
    def __init__(self, joints=None, shape=None, skin=None, verts=None):
        self.joints = joints
        self.shape = shape = skin
        self.verts = verts

Let’s use that when we parse the file, let’s then store the data we paresed in our new shape class:

#the function takes a path to the file we want to parse
def parseFile(self, path):
    root = xml.etree.ElementTree.parse(path).getroot()
    #set the header info
    for atype in root.findall('headerInfo'):
        self.fileName = atype.get('fileName')
    for atype in root.findall('weights'):
        jnt = atype.get('source')
        shape = atype.get('shape')
        clusterName = atype.get('deformer')
        if shape not in self.shapes.keys():
            self.shapes[shape] = self.skinnedShape(shape=shape, skin=clusterName, joints=[jnt])
            s = self.shapes[shape]

So now we have a dictionary of our shape classes, and each knows the shape, cluster name, and all influences. This is important because, if you read the previous post, the weights will only load onto a skinCluster with the exact same number and names of joints.
Now we write a method to apply the weight info we parsed:

def applyWeightInfo(self):
    for shape in self.shapes:
        #make a skincluster using the joints
        if cmds.objExists(shape):
            ss = self.shapes[shape]
            skinList = ss.joints
            cluster = cmds.skinCluster(, tsb=1)
            fname = self.path.split('\\')[-1]
            dir = self.path.replace(fname,'')
            cmds.deformerWeights(fname , path = dir,, im=1)

And there you go. Let’s also write a method to export/save the skinWeights from a list of meshes so we never have to use the Export DeformerWeights tool:

def saveWeightInfo(self, fpath, meshes, all=True):
    t1 = time.time()
    #get skin clusters
    meshDict = {}
    for mesh in meshes:
        sc = mel.eval('findRelatedSkinCluster '+mesh)
        #not using shape atm, mesh instead
        msh =  cmds.listRelatives(mesh, shapes=1)
        if sc != '':
            meshDict[sc] = mesh
            cmds.warning('>>>saveWeightInfo: ' + mesh + ' is not connected to a skinCluster!')
    fname = fpath.split('\\')[-1]
    dir = fpath.replace(fname,'')
    for skin in meshDict:
        cmds.deformerWeights(meshDict[skin] + '.skinWeights', path=dir, ex=1, deformer=skin)
    elapsed = time.time()-t1
    print 'Exported skinWeights for', len(meshes), 'meshes in', elapsed, 'seconds.'

You give this a folder and it’ll dump one file per skinCluster into that folder.
Here is the final class we’ve created [], and let’s give it a test run.

sdw = skinDeformerWeights()
>>>Exported skinWeights for 214 meshes in 2.433 seconds.

Let’s now load them back, we will iterate through the files in the directory and parse each, applying the weights:

import os
path = "e:\\gadget\\"
files = 0
for file in os.listdir(path):
    if file.endswith(".skinWeights"):
        fpath = path + file
        sdw = skinDeformerWeights(path=fpath)
        files += 1
elapsed = time.time() - t1
print 'Loaded skinWeights for', files, 'meshes in', elapsed, 'seconds.'
>>> Loaded skinWeights for 214 meshes in 8.432 seconds.

“>>> Loaded skinWeights for 214 meshes in 8.432 seconds.”

So that’s a simple 50 line wrapper to save and load skinWeights using the deformerWeights command. No longer do we need to write C++ API plugins to save/load weights quickly.

posted by Chris at 1:27 AM  

Saturday, September 24, 2016

DeformerWeights Command, Cloaked Savior?


This post was originally going to be entitled ‘Why Everyone Writes their Own Skin Exporter’. Maya’s smooth skin weight export tool hasn’t changed since Maya 4.0 when it was introduced 15 years ago. It saves out greyscale weightmaps in UV space, one image per joint influence per mesh. The only update they have done in 15 years is change the slider to go from a max of 1024 pixels to 4096, and now 8192!

Autodesk understands the need and importance of a skin weight exporter, they even ship it as a C++ example in their Maya Developer Kit / SDK. So why are they still shipping this abomination above?


Like blind data discussed before, in pythonland we cannot set skinweights in one go, we must iterate through the entire mesh setting skin weights one vertex at a time. This means a Maya feature or C++ plugin can save and load skin weights in seconds that take python 15 minutes.

I have written lots of skin weight save/load crap in my time, and as I sat down and started to do that very thing in an instructional blog post one night, we had an interesting discussion in the office the next day. ADSK added ‘Export Deformer Weights‘ in 2011, but it has never really worked. I don’t know a single person who uses it. But it does save and load ‘deformer’ weights via the C++ API –so there’s real promise here!



Save/load skin weights fast without writing a custom Maya plugin? This is kinda like the holy grail, which is ridiculous, but you should see the lengths people go to eek out a little more performance! My personal favorite was MacaroniKazoo back in 2010 reaching in and setting the skinCluster weight list directly by hand using an unholy conglomeration of python API and MEL commands. Tyler Thornock has a post that builds on this here.

So I asked the guys if anyone had looked at Export Deformer Weights recently, everyone either hadn’t heard of it, heard it was shit, or had some real first-hand experience of it being “A bit shit.” But still –the promise was there!

Export Deformer Weights: Broken and Backwards

So, first thing’s first, I made an awesome test case. I am going to go over all the gotchas and things that are broken, but if you don’t want to take this voyage of discovery, skip this section.


I open the deformer weight export tool, and just wow.. I mean the UI team really likes it’s space:


I save my skin weights to an XML file, delete history on the mesh, and open the import UI:


Gotcha 1) It requires a deformer to load weights onto. You need to re-skin your mesh.

I re-skinned my mesh and loaded the XML using Index.. drumroll..


Well, this is definitely not applying the weights back by vertex index. I decided to try ‘Nearest’:


Gotcha 2) Of the options [Index, Nearest, Over], ‘Index’ is somehow lossy, and anything other than ‘Index’ seems to crash often, ‘Nearest’ seems totally borked. (above)

So this was when I just began to think this was a complete waste of my time. I was pretty annoyed that they even shipped a tool like this, something that is so needed and so important, yet crashes frequently and completely trashes your data when it does work.


Not Taking ‘Broken and Backwards’ for an Answer

I am already invested and, not understanding how loading weights by point index could be lossy and broken, I decided to look at the XML file. The tool writes out one XML file per skinCluster, here’s a rundown of the file format:

Mesh Info (Shape) – the vertices of the shape are stored local space x,y,z and corresponding index

<?xml version="1.0"?>
  <headerInfo fileName="C:/Users/chris.evans/Desktop/test_sphere.xml" worldMatrix="1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 "/>
  <shape name="pSphereShape1" group="7" stride="3" size="382" max="382">
    <point index="0" value=" 0.148778 -0.987688 -0.048341"/>
    <point index="1" value=" 0.126558 -0.987688 -0.091950"/>
    <point index="2" value=" 0.091950 -0.987688 -0.126558"/>

Joints (Weights) – There is one block per joint that calls out each vertex that it have influences for on the shape

  <weights deformer="skinCluster1" source="root" shape="pSphereShape1" layer="0" defaultValue="0.000" size="201" max="380">
    <point index="0" value="0.503"/>
    <point index="1" value="0.503"/>
    <point index="2" value="0.503"/>

And that’s it, not a lot of data, nothing about the skinCluster attributes or options, no support for spaces like UV or world. (odd, since it’s had UV support for 15 years)

Next I decided to run the tool again and see what command it was calling, I then looked up the command documentation and here’s where it gets interesting, go ahead, take a look!


So now I am hooked, someone is putting some thought into this –at least on some level.


I don’t at all understand why the UI has none of these options, but I need to get this working. If you read through the docs, the command also supports:

  • Exporting multiple skinClusters/shapes/deformers per XML file
  • Exporting skinCluster/deformer attributes like ‘skinningMethod’ and ‘envelope’
  • Local and world space positions with a positional tolerance

“Someone is putting some thought into this”

So I started trying to figure out why a file format that explicitly knows every influence of every vertex by index and inf name, doesn’t load weights properly. After some trials I hit gotcha #3:

Gotcha 3) Of the options the weights only load properly if the skinCluster has the *exact* same influences it was saved with. Which really makes no sense, because the file format has the name of every joint in the old skinCluster.

So now I had it working, time to wrap it and make it useful.

The Documentation is a Lie.

So, first thing’s first, I did a speed test.

Importing with deformerWeights was about 125 times faster: Gold mine.

But I just couldn’t get some of the flags to work, I thought I was just a moron, until I finally tried the code example in the ADSK Maya documentation, which FAILS. Let’s first look at the -vc flag, which is required to load using ‘bilinear’ or ‘barycentric’ mapping/extrapolation:

cmds.deformerWeights ("testWeights.xml", ex=True, vc=True, deformer="skinCluster1")
# Error: Invalid flag 'vc'
# Traceback (most recent call last):
#   File "<maya console>", line 1, in <module>
# TypeError: Invalid flag 'vc' #

Gotcha 4) The python examples do not work. -vertexConnections flag doesn’t work, -attribute flag doesn’t work, so no saving skincluster metadata like ‘skinningMethod’, etc. Because of that, ‘barycentric’ and other methods that need vertex connection info do not work. The ‘deformer’ flag shows that it takes a list of deformers and writes them all to one file, but this is not true, it takes a single string name of a deformer.

I now know why the UI doesn’t have all these cool options! –they don’t work!

Gotcha 5) It doesn’t take a file path, to save a file to a path you need to specify the filename, and then the path separate.

cmds.deformerWeights ("testWeights.xml", path='d:\\myWeight\\export\\folder\\', ex=True, deformer="skinCluster1")

Perhaps someone fixed this stuff, documented it, and then reverted the fix, but this has been around since 2011.. I tried the above in maya 2016 latest service pack and all my links above are to that version of the documentation.

I wasn’t really intending to write this much, so now that we know this can import weights 125 times faster, we’ll make a tool to utilize it. Stay tuned!

posted by Chris at 2:15 AM  

Tuesday, June 7, 2016

Displaying half the data: But twice as fast!


In Maya 2016, by default, Node Editor cannot display more than 500 connections. Let’s put this in perspective. Let’s say you have two attributes and each drives 20 joints. BOOM, you’re over the limit bro.. what you doin? Why you breakin’ Maya?

The solution, the warning says is to set an option var. Oh yeah that’s a frickin’ charm, that’s like unlocking all streets on my GPS by loosening a nut on the underside of my car.

I count on this visual editor to display connections, that’s all it needs to do. I have had rigs that make this update once every 10 seconds, but at least I could find what I was looking for and troubleshoot it. I would much rather a drop in fidelity, like stop drawing nice curves or something –than just randomly not displaying data.

How is that even an option? Sometimes my mind wanders:

“Sir, we noticed that on some systems, the software can get slow if we display more than 500 connections.”
“How did you fix it?”
“Well we just decided not to show more than 500 connections.”
“Stellar work Raymond. Can this limit be adjusted in the Settings and Preferences? Maybe the first time Maya opens or the user opens Node Editor it can be set based on his hardware.”
“Well right now it’s an optionVar, I mean that works well enough for me.”

Here’s what you can enter into the mel evaluation line in the lower left to set your max display depth to ten million:

optionVar -iv "editorConnectionScanLimit" 10000000;

I think ADSK probably hates me by now. ¬.¬

posted by Chris at 12:11 AM  

Tuesday, April 19, 2016

ADSK Ships Pose Space Deformer, Ignores Clavicle

Oh the clavicle.. you’re like the bastard redheaded stepchild.. is this guy an animator or what?

Anyway, sixteen years after the publication of Pose Space Deformation: A Unified Approach to Shape Interpolation and Skeleton-Driven Deformation, and fourteen years after Michael Comet released the poseDeformer plugin for Maya, which became the industry standard: Autodesk has now implemented the feature into the software.

In what seems like my early 20’s being acted out before my eyes, ‘Combination Sculpting’, the way of associating a fixer shape when other shapes are in a certain configuration (pioneered by Bay Raitt on Gollum in 2002), is also not a feature that will ship in Maya.

Riggers rejoice! We have voted this up on all feedback forums for years, and ADSK created the tool with direct involvement from the rigging community, great that it can finally see daylight. Now, all you Extension2 users grab it and work out the remaining kinks before everyone gets it in Maya 2017!

And if you’re upgrading to EXT2 today, don’t forget, FBX anim export/bake is fixed!

posted by Chris at 3:41 PM  

Monday, April 4, 2016

Off on a Tangent

Plutarch tells that Alexander the Great visited the Libyan Sibyl, and was given the correct constellation of checkboxes and steps to get meshes from Max to Maya properly.

Plutarch tells that Alexander the Great visited the Libyan Sibyl in search of the correct constellation of checkboxes and steps to get meshes from Max to Maya.

I have spent many years of my life in studios where characters are modeled in a package other than Maya (often Max) and imported into Maya via FBX. Having worked along side great character artists like Hanno Hagedorn, Abdenour Bachir, and most recently Kevin Lanning and his team here at Epic, I cannot tell you how many hours of our lives were devoted to trying to get mesh tangents into the final product that resembled what they were in the original sculpt/bake. Not to mention brilliant pipeline programmers like Bogdan Coroi, or James Goulding‘s team here at Epic, many hours have been spent trying to solve this issue.

Sometimes it seemed some mystical channeling whereby some constellation of export or import checkboxes along with maybe layering an edit mesh modifier on top of your character before export to Maya worked. Sometimes the solution seemed to have been exporting only triangulated meshes to Maya, whereby you needed a (fragile) pipeline to allow you to have quaded for skinning and triangulated from Max for export.

Well, as it turns out, Maya has always ignored all mesh tangent data on FBX import.

I hope this post saves you some headache. At Crytek we looked to change the pipeline to store all normal maps in world space, another, more pragmatic solution, proposed by Jeremy Ernst here at Epic is to give the engine the static mesh from Max and the skinned mesh from Maya and just transfer the original data. Scott Parrish told me that his team bakes against the skinned FBX as it comes out of UE4, this is another way of solving the issue.

I also understand this is not a simple issue, all DCC packages work differently. Max allows users to turn edges, etc.. But it’s good to know that we’re also not crazy.

posted by Chris at 9:46 PM  

Thursday, March 31, 2016

Why Does Everyone Write Their Own FBX Exporter?


When it comes to characters or character animation, any studio doing anything remotely complex has to write their own FBX exporter. But why?


It’s true that the FBX export options dialog is so convoluted that even ADSK has created a front end for exporting FBX to Unreal or Unity, but the main reason boils down to one issue: Maya FBX Export does not bake animation properly.

I *WANT* to use the FBX Exporter core (from Python). As a technical artist, I want a C++ plugin efficiently traversing the dag and baking/exporting animation. I do not want to dupe a skeleton and walk the timeline using python, calling scene update on the entire DAG every frame. The FBX Exporter doesn’t even have python exposure, so we have to eval MEL commands; it’s just not ideal to say the least.

Where Have the Attrs Gone?

When exporting characters elsewhere, you want to use FBX to get important data there as well. Not just an animated transform. In this test file [fbx_test], you see a sphere, skinned to a joint. That joint has an attribute called important_data. In games especially, we can use this scalar to drive a material parameter such as a wrinkle on a face, a blendshape, to allow an animator to blend off a cloth solver, anything really. So here’s our important data, driven my a multiply-divide node:


Export this scene as an FBX, being sure to tell it to bake complex animation:

mel.eval("FBXExportBakeComplexAnimation -v true")

Now import it back into an empty Maya scene, your important_data is gone!

To make matters worse, in the FBX ascii file I see an anim curve for my important data:

;AnimCurveNode::important_data, Model::joint1
C: “OP”,1519473056,205108448, “important_data”

So I am at a loss as to why programs using the FBX SDK do not get this data on import, including Maya.

DISCLAIMER: On multiple occasions I have reached out to ADSK asking them to please address this issue, but it has not happened. From the image above (new game export options), you can see that, on some level, ADSK wants to make this experience better, so please echo me in asking them to fix this issue.

In line with the title of this post, I have written, and would gladly give out an exporter for UE4, but as it uses the C++ bake / FBX Exporter from Maya, I would prefer to wait until this gets fixed. The ‘Bake Root’ option was actually created by Kevin Vassey at Epic because UE4 imports custom attrs on the root joint, so we just bake that if needed. This is a little puny exporter; internally we use Jeremy’s A.R.T. Toolkit for exporting animation in production.



UPDATE 1: Thanks for the shares and responses. Some of you have told me that you have also talked to ADSK about this, and others say it was fixed at one time, but is now broken. ADSK has reached out in the comments and has created a bug number in their tracking system. I will update later on it’s priority/timeframe if I get more info!

UPDATE 2: I can verify, this is fixed if you grab FBX 2016.1.2 or later from here. It should be fixed in future versions of Maya, and the bug was fixed before I wrote this post, but was not in Maya 2016. In the comments, Chris Dardis said that the fix was available in a non public cut of Maya he was using. As the link is not compiled, I have uploaded a compiled Maya 2016 FBX 2016.1.2 plugin here.

posted by Chris at 10:01 PM  

Friday, January 23, 2015

BEWARE: Maya 2015 EXT / 2014 Constraint Incompatibility


I like to live on the edge when it comes to new versions of Maya, but rarely does it cause a major issue for me. A wise man once said ‘the moment you stop respecting something is when it bites you in the ass’. Well consider me bitten.

Anytime a rig I created with 2015 is loaded in 2014 it is irreparably broken. Spewing tons of errors like this one:
// Error: file: C:/Users/chris.evans/Desktop/ line 210: The orientConstraint ‘nurbsSphere1_orientConstraint1’ has no ‘w0’ attribute. //

Diffing two constrained spheres, I noticed that there’s a new flag called ‘DCB’ or Disconnect Behavior.  Batch delete that and it seems that 2015 rigs can work in 2014.

But as usual, be a better person than me and just don’t open a file with a version of Maya your animators aren’t on. 😀

posted by Chris at 10:57 PM  

Friday, October 17, 2014

Embedding Icons and Images In Python with XPM


As technically-inclined artists, we often like to create polished UIs, but we have to balance this with not wanting to complicate the user experience (fewer files the better). I tend to not use too many icons with my tools, and my Maya tools often just steal existing icons from Maya: because I know they will be there.

However, you can embed bitmap icons into your PySide/PyQt apps by using the XPM file format, and storing it as a string array. Often I will just place images at the bottom of the file, this keeps icons inside your actual tool, and you don’t need to distribute multiple files or link to external resources.

Here’s an example XPM file:

/* XPM */
static char *heart_xpm[] = {
/* width height num_colors chars_per_pixel */
"7 6 2 1",
/* colors */
"` c None",
". c #e2385a",
/* pixels */

This is a small heart, you can actually see it, in the header you se the ‘.’ maps to pink, you can see the ‘.’ pattern of a heart. The XPM format is like C, the comments tell you what each block does.
Here’s an example in PySide that generates the above button with heart icon:

import sys
from PySide import QtGui, QtCore
def main():
    app = QtGui.QApplication(sys.argv)
    heartXPM = ['7 6 2 1','N c None','. c #e2385a','N..N..N',\
    w = QtGui.QWidget()
    w.setWindowTitle('XPM Test')
    w.button = QtGui.QPushButton('XPM Bitmaps', w)
if __name__ == '__main__':

You need to feed QT a string array, and strip everything out. Gimp can save XPM, but you can also load an image into xnView and save as XPM.

Addendum: Robert pointed out in the comments that pyrcc4, a tool that ships with PyQt4, can compile .qrc files into python files that can be imported. I haven’t tried, but if they can be imported, and are .py files, I guess they can be embedded as well. He also mentioned base64 encoding bitmap images into strings and parsing them. Both these solutions could really make your files much larger than XPM though.

posted by Chris at 3:23 PM  

Thursday, October 16, 2014

Tracing and Visualizing Driven-Key Relationships


Before I get into the collosal mess that is setting driven keys in Maya, let me start off by saying, when I first made an ‘SDK’ in college, back in 1999, never did I think I would still be rigging like this 15 years later. (Nor did I know that I was setting a ‘driven key’, or a ‘DK’ which sounds less glamorous)

How The Mess is Made


Grab this sample scene [driven_test]. In the file, a single locator with two attributes drives the translation and rotation of two other locators. Can’t get much ismpler than that, but look at this spaghetti mess above! This simple driven relationship created 24 curves, 12 blendWeighted nodes, and 18 unitConversion nodes. So let’s start to take apart this mess. When you set a driven-key relationship, it uses an input and a curve to drive another attribute:


When you have multiple attributes driving a node, maya creates ‘blendWeighted’ nodes, this blends the driven inputs to one scalar output, as you see with the translateX below:


Blending scalars is fairly straight forward, but for rotations, get ready for this craziness: A blendWeighted node cannot take the output of an animCurveUA (angle output), the value must first be converted to radians, then blended. But the final node cannot take radians, so the result must be converted back to an euler angle. This happens for each channel.


If you think this is retarded; welcome to the club. It is a very cool usage of general purpose nodes in Maya, but you wouldn’t think so much of rigging was built on that, would you? That when you create a driven-key it basically makes a bunch of nodes and doesn’t track an actual relationship, because of this, you can’t even reload a relationship into the SDK tool later to edit! (unless you traverse the spaghetti or takes notes or something)

I am in love with Node Edtor, but by default hypergraph did hide some of the spaghetti, it used to hide unitConversions as ‘auxiliary nodes’:


Node Editor shows unitConversions regardless of whether aux nodes are enabled, I would rather see too much and know what’s going on than have things hidden, but maybe that’s just me. You can actually define what nodes are considered aux nodes and whether ‘editors’ show them, but I am way off on a tangent here.

So just go in there and delete the unit conversion yourself and wire the euler angle output, which you would think is a float.. into the blendWeighted input, which takes floats. Maya won’t let you, it creates the unitConversion for you because animCurveUA outputs angles, not floats.

This is why our very simple example file generated over 50 nodes. On Ryse, Maruis’ face has about 73,000 nodes of driven-key spaghetti. (47,382 curves,  1,420 blendWeighted, 24,074 unitConversion)

Finding and traversing

So how do we find this stuff and maybe query and deal with it? Does Maya have some built in way of querying a driven relationship? No. Not that I have ever found. You must become one with the spaghetti! setDrivenKeyframe has a query mode, but that only tells you what ‘driver’ or ‘driven’ nodes/attrs are in the SDK window if it’s open, they don’t query driver or driven relationships!

We know that these intermediate nodes that store the keys are curves, but they’re a special kind of curve, one that doesn’t take time as an input. Here’s how to search for those:

print len("animCurveUL","animCurveUU","animCurveUA","animCurveUT")))

So what are these nodes? I mentioned above that animCurveUA puts out an angle:

  • animCurveUU – curve that takes a double precision float and has a double output
  • animCurveUA – takes a double and outputs an angle
  • animCurveUL – takes a double and outputs a distance (length)
  • animCurveUT – takes a double and outputs a time

When working with lots of driven-key relationships you frequently want to know what is driving what, and this can be very difficult because of all the intermediate node-spaghetti. Here’s what you often want to know:

  • What attribute is driving what – for instance, select all nodes an attr drives, so that you can add them back to the SDK tool. ugh.
  • What is being driven by an attribute

I wrote a small script/hack to query these relationships, you can grab it here [drivenKeyVisualizer]. Seriously, this is just a UI front end to a 100 line script snippet, don’t let the awesomeness of QT fool you.


The way I decided to go about it was:

  1. Find the driven-key curves
  2. Create a tiny curve class to store little ‘sdk’ objects
  3. List incoming connections (listConnections) to get the driving attr
  4. Get the future DG history as a list and reverse it (listHistory(future=1).reverse())
  5. Walk the reversed history until I hit a unitConversion or blendWeighted node
  6. Get it’s outgoing connection (listConnections) to find the plug that it’s driving
  7. Store all this as my sdk objects
  8. Loop across all my objects and generate the QTreeWidget

Here’s how I traversed that future history (in the file above):

 #search down the dag for all future nodes
 futureNodes = [node for node in cmds.listHistory(c, future=1, ac=1)]
 #reverse the list order so that you get farthest first
 drivenAttr = None
 #walk the list until you find either a unitConversion, blendWeighted, or nothing
 for node in futureNodes:
     if cmds.nodeType(node) == 'unitConversion':
             drivenAttr = cmds.listConnections(node + '.output', p=1)[0]
             cmds.warning('blendWeighted node with no output: ' + node)
     elif cmds.nodeType(node) == 'blendWeighted':
             drivenAttr = cmds.listConnections(node + '.output', p=1)[0]
             cmds.warning('blendWeighted node with no output: ' + node)
 if not drivenAttr:
     drivenAttr = cmds.listConnections(c + '.output', p=1)[0]
 if drivenAttr:
     dk.drivenAttr = drivenAttr
     cmds.warning('No driven attr found for ' + c)

This of course won’t work if you have anything like ‘contextual rigging’ that checks the value of an attr and then uses it to blend between two banks of driven-keys, but because this is all general DG nodes, as soon as you enter nodes by hand, it’s no longer really a vanilla driven-key relationship that’s been set.

If you have a better idea, let me know, this above is just a way I have done it that has been robust, but again, I mainly drive transforms.

 What can you do?

Prune Unused Pasta


Pruned version of the DAG

By definition, when rigging something with many driven transforms like a face, you are creating driven-key relationships based on what you MIGHT need. This goes for when making the initial relationship, or in the future when you maybe want to add detail. WILL I NEED to maybe translate an eyelid xform to get a driven pose I want.. MAYBE.. so you find yourself keying rot/trans on *EVERYTHING*. That’s what I did in my example, and the way the Maya SDK tool works, you can’t really choose which attrs per driven node you want to drive, so best to ‘go hunting with a shotgun’ as we say. (shoot all the trees and hope something falls out)

Ok so let’s write some code to identify and prune driven relationships we didn’t use.

CAUTION: I would only ever do this in a ‘publish’ step, where you take your final rig and delete crap to make it faster (or break it) for animators. Also, I have never used this exact code below in production, I just created it while making this blog post as an example. I have run it on some of our production rigs and haven’t noticed anything terrible, but I also haven’t really looked. 😀

def deleteUnusedDriverCurves():
    for driverCurve in"animCurveUL","animCurveUU","animCurveUA","animCurveUT")):
        #delete unused driven curves
        if not [item for item in cmds.keyframe(driverCurve, valueChange=1, q=1) if abs(item) &gt; 0.0]:

This deletes any curves that do not have a change in value. You could have multiple keys, but if there’s no curve, let’s get rid of it. Now that we have deleted some curves, we have some blendWeighted nodes that now aren’t blending anything and unitConversions that are worthless creatures. Let’s take care of them:

def deleteUnusedBlendNodes():
    #rewire blend nodes that aren't blending anything
    for blendNode in'blendWeighted'):
        if len(cmds.listConnections(blendNode, destination=0)) == 1:
            curveOutput = None
            drivenInput = None
            #leap over the unitConversion if it exists and find the driving curve
            curveOutput = cmds.listConnections(blendNode, destination=0, plugs=1, skipConversionNodes=1)[0]
            #leap over the downstream unitConversion if it exists
            conns = cmds.listConnections(blendNode, source=0, plugs=1, skipConversionNodes=1)
            for conn in conns:
                if cmds.nodeType(conn.split('.')[0]) == 'hyperLayout': conns.pop(conns.index(conn))
            if len(conns) == 1:
                drivenInput = conns[0]
                cmds.warning('BlendWeighted had more than two outputs? Node: ' + blendNode)
            #connect the values, delete the blendWeighted
            cmds.connectAttr(curveOutput, drivenInput, force=1)
            print 'Removed', blendNode, 'and wired', curveOutput, 'to', drivenInput

We find all blendWeighted nodes with only one input, then traverse out from them and directly connect whatever it was the node was blending, then we delete it. This is a bit tricky and I still think I may have missed something because I wrote this example at 2am, but I ran it on some rigs and haven’t seen an issue.

Here are the results running this code on an example rig:


Create a Tool To Track / Mirror / Select Driven Relationships

This is a prototype I was working on at home but shelved, I would like to pick it up again when I have some time, or would be open to tossing it on github if people would like to help. It’s not rocket science, it’s besically what Maya should have by default. You just want to track the relationships you make, and also select all nodes driven by an attr. Also mirror their transforms across an axis (more geared toward driven transforms).


Write Your Own Driver Node

Many places create their own ‘driven node’ that just stores driven relationships. Judd Simantov showed a Maya node that was used on Uncharted2 to store all facial poses/driven relationships:


The benefits of making your own node are not just in DAG readability, check out the time spent evaluating all these unitConversion and blendWeighted nodes in a complex facial rig (using the new Maya 2015 EXT 1 Profiler tool) –that’s over 760ms! (click to enlarge)


Though it’s not enough to make a node like this, you need to make a front end to manage it, here’s the front end for this node:


Give Autodesk Feedback!


As the PSD request is ‘under review’, maybe when they make the driver, they can make a more general framework to drive things in Maya.


As you can see, there are many ways to try to manage and contain the mess generated by creating driven-key relationships.  I would like to update this based on feedback, it’s definitely not an overview, and speaks more to people who have been creating driven-key relationships for quite some time.  Also, if you would find a tool like my Maya SDK replacement useful, let me know, especially if you would like to help iron it out and/or test it.

posted by Chris at 3:16 PM  

Wednesday, October 15, 2014

Reminder: Maya saves ScriptEditor tabs on crash


This is just a reminder that for some time now Maya has been saving all ScriptEditor tabs on crash. I frequently bump into people who don’t know or don’t remember this. If your Maya says that it’s attempting to save in your /Temp folder, it’s also saving all your ScriptEditor tabs.

posted by Chris at 11:56 AM  

Tuesday, September 30, 2014

Why Storing String Refs is Dangerous and Irresponsible

Yes! Definitely a clickbait title if there ever was one; but this is important! Some of you contacted me regarding my earlier post (Don’t Use String Paths), where I said you should never use string names to keep track of anything even vaguely important in Maya.

This problem is so fundamental in Maya that the initial Python code test I came up with for the Crytek Technical Art Dept used to be a simple maya node renamer. No joke.


From time to time I see code that attempts reach out into the DAG and grab nodes by concatenating tokens like:

character_side + '_' + character_name + '_arm_' + UID + '_' + type

This is SUPER dangerous. You are basically guessing that an object exists, and betting your whole tool or codebase on this, it’s super fragile.

Let’s use the example of making a turtle, let’s call him ‘Red’, he’s red, and his shell is soft:

cmds.joint(name = 'red' + '_' + shell_hardness + '_' + animal_type + '_hand')'red' + '_' + shell_hardness + '_' + animal_type + '_hand')
# Error: No object matches name: red_soft_turtle_hand
# Traceback (most recent call last):
#   File "", line 1, in 
# ValueError: No object matches name: red_soft_turtle_hand #

First up, you should *never* do this, when you are creating a node with a name you are making from scratch, you *must* first check if it exists. If it exists, Maya will alter the name and you will not be able to find it, your code will immediately break either with the above error, or this one:'red_soft_turtle_hand')
# Error: More than one object matches name: red_soft_turtle_hand
# Traceback (most recent call last):
#   File "", line 1, in 
# ValueError: More than one object matches name: red_soft_turtle_hand #


So here’s something a bit more safe, but still not recommended in situations where this is important:

hand_jnt = cmds.joint(name = 'red' + '_' + shell_hardness + '_' + animal_type + '_hand')
print hand_jnt
# >>: red_soft_turtle_hand1

Here, we’re storing the node created in a variable, if the name is not the name we expected, we still know how to find the node. This is a lot safer, your code will continue to run, but it’s also not a great way of working. Maya created our joint, but called it ‘red_soft_turtle_hand1’, but, because we stored the node returned from the command creating it into a variable, we can print hand_jnt, and it will return ‘red_soft_turtle_hand1’.

To be super clear, this is more safe because:

  • Maya is resolving any name clashes for you automatically
  • You get the *real* name returned to you from Maya upon node creation
  • If you create a node name that exists elsewhere, it returns you a full path automatically

This is somewhat acceptable in code that’s building a rig. Wham! Bam! You created a node and then did something with it seconds later and never looked back!


So let’s say you’re doing the slightly safer way, and you’re storing long paths of nodes which Maya gives you the name of. As soon as you store a long path, it’s stale.  This is what I meant above, the longer you store this information, the less reliable it is.

Let’s look at our joint, its long path is:


If *any* of the six parent nodes names change, just a single character! You’re dead in the water. If the hierarchy changes: you’re dead in the water.

If you’re working with someone who defends the above by saying the following

  • “Oh, the hierarchy will never change”
  • “Oh, there will never be a node with the same name”
  • “Oh, no one will ever rename any of these”

The Maya DAG is the wild west.
You WILL have duplicate node names.
You WILL have hierarchy changes.
Professional tools don’t only work with a bunch of caveats,
if it’s a valid Maya scene, your code should work with it:


There are clear ways to overcome the challenge above, they’re not secrets, they are actually the professional way Autodesk tells you to do it in the docs, code examples, and their own implementations.

You don’t *have* to work 100% in the API or PyMel, but like we said before, as soon as possible get a pointer to your object. Think of any string path as having a short half-life.

Using The Maya Python API

In C++ when you pass around an object, you have a pointer to it’s location in memory. The name and DAG path can change, but you can at any time get the current name, and path. Fresh, not stale!

import maya.OpenMaya as om
m_dg = om.MFnDagNode()
m_obj = m_dg.create("joint")
m_dagPath = om.MDagPath()
if m_obj.hasFn(om.MFn.kDagNode):
    om.MDagPath.getAPathTo( m_obj, m_dagPath )
    print m_dagPath.fullPathName()
# >>: |joint1

Using PyMel

PyMel wraps the API, a PyMel object is a *real* python object that stores the actual MObject, but you don’t have to use the API.

myJoint = joint(name='doesnt_even_matter')
print myJoint.fullPath()
# >>: |root|pelvis|shell|some_other_joint|arm1|arm2|red_soft_turtle_hand|doesnt_even_matter
#In case you never used PyMel, here's some examples of convenience functions:

Notice that PyMel is handing you back python objects, you can ask the object for it’s full path at any time, and it will be fresh data.

Using Message Attrs

Message attributes are how Alias decided users will serialize relationships between nodes. I have removed this section because future-me wrote an entire post dedicated to this (The Mighty Message Attribute).


The important thing here is that you decide a consistent way of using the above. At Crytek we tried to have all core functions that accepted or passed DG nodes, do the handoff as MObjects. We serialized all important relationships with message attrs, and when there was too much data for that to be efficient, we stamped nodes with metadata.


If you have an existing pipeline where you build a lot of string references by concatenating tokens like the example above, you can make a convenience function that will validate your assumptions and deal gracefully with issues. Something like my_utils.get_node(‘my’ + ‘_’ + ‘thing’)

posted by Chris at 3:28 PM  

Wednesday, September 24, 2014

Maya DAG nodes in QT tools (don’t use string paths)



String paths to Maya objects are _fragile_. This data is stale as soon as you receive it, long paths are better than short, and taking into account namespaces is even better –BUT when it comes down to it, the data is just old as soon as you get it, and there is a better way.

A Full Path that is Never Stale

If you store a node in a Maya Python API MDagPath object, you can ask at any time and get its full path. Because it’s basically a pointer to the object in memory. Here’s an example:

import maya.OpenMaya as om
#create and populate an MSelectionList
sel = om.MSelectionList()
#create and fill an MDagPath
mdag = om.MDagPath()
sel.getDagPath(0, mdag)
#At any time request the current fullpath to the node
print mdag.fullPathName()

CHALLENGE: Can you write a renaming tool that doesn’t use string path node representation? That works with all DG nodes? Without using PyMel? 😉

Embedding Maya Python Objects in PyQT (UserRole)

So.. if we can store a pointer to an object, how do we track that with QT UIs, pass it around, etc? It can be really difficult to find a way to be responsible here, you have all these lists and node representations, and in the past I would try to sync lists of python MDagPath objects, or make a dictionary that mirrors the listView… but there is a way to store arbitrary python objects on list and tree widgetitems!

In this code snipet, I traverse a selection and add all selected nodes to a list, I add ‘UserRole’ data to each and set that data to be an MDagPath:

    listWids = []
    for item in nodes:
        sel = om.MSelectionList()
        mdag = om.MDagPath()
        sel.getDagPath(0, mdag)
        name = mdag.fullPathName()
        if not self.longNamesCHK.isChecked():
            name = name.split('|')[-1]
        wid = QtGui.QListWidgetItem(listWid)
        wid.setData(QtCore.Qt.UserRole, mdag)

Now, later when we want to get the full path name of this widgetItem, we just ask it for this data. That returns the MDagPath object, and we can ask the object for the current full path to the node:

print self.drivenNodeLST.currentItem().data(QtCore.Qt.UserRole).fullPathName()

So this is a good way to have arbitrary data that travels wit the QT node description, which is usually some kind of widget.

posted by Chris at 12:40 PM  

Thursday, September 18, 2014

Ryse SIGGRAPH 2014 Reel

I mentioned that we won “best Realtime Graphics” at this year’s SIGGRAPH conference, but I never linked the video:

posted by Chris at 4:24 PM  

Tuesday, August 26, 2014

Multi-Resolution Facial Rigging

At SIGGRAPH we discussed a bit about our facial pipeline that we haven’t talked about before. Namely, facial LODs and multi-platform facial rigging.

I would like to start by saying that we spent a _LOT_ of time thinking about facial levels of detail on Ryse, and put a lot of effort into the area. I know this is a long post, but it’s an important one.


Lowest Common Denominator

As the ‘next-generation’ seems to be largely defined by mult-platform titles it seems valuable to focus on ways to increase fidelity on next generation hardware while still being able to target older hardware specs. That said, I have yet to see a pipeline to do this. Most next gen games have skeletons and animations limited by the lowest common denominator of the last generation, often Playstation 3.

When you wonder why your awesome next gen game doesn’t seem to have character models and animation like next-gen only titles, this is why.

It’s very easy to increase texture resolution by having a pipeline where you author high and bake to lower targets.  It’s more complicated to author meshes high and publish to lower targets, we did this on Crysis 1 and 2, High end PC saw higher mesh resolution characters than Xbox 360. I would say it’s the hardest to make rigs, deformers, and animations for a high spec hardware target and create a process to publish lower fidelity versions. No one wants to have different character skeletons on each hardware platform.


You Deserve an Explanation

When we released the specs of our faces, people understandably were a bit taken aback.  Why on earth would you need 250 blendshapes if you have 260 joints? This above is actually a slide from our asset creation SIGGRAPH course that resonated very well with the audience.

Let’s take a look at some goals:

  1. Cut-scene fidelity in gameplay at any time- no cut-scene rigs
  2. Up to 70 characters on screen
  3. Able to run on multiple hardware specs

The only way to achieve the first two is through a very aggressive and granular level of detail (LOD) scheme. Once having that LOD system in place, the third item will come for free, as it did on our previous titles. However, as we have LODed meshes and materials, we had never LODed rigs.

On a feature film, we wouldn’t use joints, we would have a largely blendshape-only face.

But this doesn’t LOD well, we need to be able to strip out facial complexity in the distance and on other platforms.

Facial Level of Detail

So to achieve these goals, we must aggressively LOD our character faces.

Let’s generate some new goals:

  • Improve LOD system to allow the swapping or culling of skinned meshes per-mesh, each at hand-tailored distances per-character instance
  • Not only swap meshes, but skinning algorithms, materials, cull blendshapes, etc..
  • One skeleton – all levels of detail stored in one nested hierarchy, disable/reveal joints at different LOD levels, as I mention above, no one wants multiple skeletons
  • One animation set – drives all layers of detail because it’s one hierarchy, only the enabled joints receive animation
  • All facial animations shareable between characters
  • Faces snapped onto bodies at runtime – “Cry parent constraint” of sorts snaps head, neck, spine4, clavs, and upper arms of facial rig to body, allowing dynamic LODing of face irrespective of body.


One Hierarchy to Rule them All

Before going into the meshes, skinning algorithms, culling, etc.. it’s important to understand the hierarchy of the face. At any given mesh LOD level, there are many joints that are not skinned. Above you see three layers of joints, 9 LOD0, 3 LOD1, and 1 LOD2.

To use a single hierarchy, but have it drive meshes at different levels, you need to accomplish the following:

  • Make sure you have three layers that can drive different facial LODs, we had something like 260/70/15 on Ryse.
  • Each layer must be driven, and able to deform that LOD alone. Meaning when creating rig logic, you must start from the highest LOD and move down the chain. The LOD0 joints above would only be responsible for skinning the details of the face at LOD0, their gross movement comes from their parent, LOD1.

Here you can see the Marius example video from our slides. Notice the ORANGE joints are responsible for gross movement and the YELLOW or GREEN leaf joints just add detail.



Why blendshapes? Isn’t 260 joints enough?

The facial hierarchy and rig is consistent between all characters. The rig logic that drives those joints is changed and tweaked, the skinning is tweaked, but no two faces are identical. the blendshapes serve two main purposes:

1) Get the joint rig back on scan. Whatever the delta from the joint rig to the scan data that is associated with that solved pose from the headcam data, bridge that. This means fat around Nero’s neck, bags under his eyes, his eyebrow region, etc.

2) Add volume where it’s lost due to joint skinning. Areas like the lips, the cheeks, rig functions like lips together, sticky lips, etc, require blendshapes.


Look at the image above, there just aren’t enough joints in the brow to catch that micro-expression on Nero’s face. It comes through with the blendshape, and he goes from looking like you kicked his dog, to his accusatory surprise when he figures out that you are Damoclese.

A Look Under the Hood: Ryse Facial LODing

Thanks to the hard work of graphics engineer Jerome Charles we were able to granularly LOD our faces. These values are from your buddy Vitallion, who was a hero and could be a bit less aggressive. Many of the barbarians you fight en masse blew through all their blendshapes in 2m not 4m.

Assets / Technologies (LOD)
CPU skinning, 8 inf, 260 joints, 230 blendshapes, tangent update, 5k  tris across multiple meshes 0-4m
CPU skinning, 8 inf, 260 joints, 3-5k across multiple meshes with small face parts culled 4-7m
GPU skinning, 4 inf, 70 joints, 2k mesh with integrated eyes 7-10m
GPU skinning , 4 inf, <10 joints, <1k mesh 10m+


Here’s a different table showing the face mesh parts that we culled and when:

Distance Face parts
4m Eyebrow meshes replaced, baked into facial texture
3m Eyelash geometry culled
3m Eye AO ‘overlay’ layer culled
4m Eye balls removed, replaced with baked in eyes in head mesh
2m Eye ‘water’ miniscus culled
3m Eye tearduct culled
3m Teeth swapped for built-in mesh
3m Tongue swapped for built-in mesh

Why isn’t this standard?

Because it’s very difficult and very complicated, there aren’t so many people in that can pull something like this off. On Ryse we partnered with my friend Vlad at 3Lateral, after 4 months working on the initial Marius facial prototype, he and his team were able to deliver 23 more facial rigs at the same fidelity in just under three months!

But also, there’s the whole discussion about whether the time and effort spent on that last 5% really pays off in the end. Why not just use PS3 facial rigs on all platforms and spend a little more on marketing? It’s happening! And those guys probably aren’t going bankrupt any time soon..  ¬.¬

I am insanely proud of what the team accomplished on Ryse. Facial rigging is nothing without a great bunch of artists, programmers, animators, etc. Here’s some good moments where the performances really come through, these are all the in-game meshes and rigs:

DISCLAIMER: All of the info above and more is publicly available in our SIGGRAPH 2014 course notes.

posted by Chris at 4:40 AM  

Monday, August 25, 2014

Maya 2015: Poly Combine Skinned Meshes?

At Crytek, we have a plugin to preserve skinning on hacking up and uniting meshes, based on this old post here. It’s released in the Tools folder of CryENGINE if you grabbed the engine on Steam. Imagine my surprise when I saw this option in Maya 2015:


It fires off a new polyUnite command called polyUniteSkinned [Maya 2015 Docs], which can merge skinned meshes? Has anyone gotten this to work? It doesn’t seem to work properly through the UI, passes ‘name’ as a flag and fails as they removed that flag. (seriously) But it seems to work in simple situations, as shown here, it didn’t work attaching a face to a body, but at least it shows ADSK is moving in the right direction!


posted by Chris at 12:42 AM  

Tuesday, July 15, 2014

RigPorn: The Last of Us

I realize most of you have seen this, but for those of you who haven’t, Judd walks people through TLOU rigs with a focus on facial as well. Really great stuff.

posted by Chris at 3:15 PM  

Thursday, July 10, 2014

RigPorn: Call of Duty: Advanced Warfare

[Click to enlarge images]

(All images taken from recent CoD marketing materials)


Here you can see the first-person hands rig, complete with camera frustum tools, and animation controls.


Close-up of the generic male rig, no face rig loaded at the moment, but still interesting.


Here’s a great shot of their first-person-hands-picker. I always love seeing how animators want to work, I really never have worked on a team who wanted a picker, much less something like this, but it’s great to see.


Another picker, maybe the full body one, or cinematics only.

Shooting at giant’s new studio in manhattan beach?


Surprisingly, this looks like it is being shot at Giant’s Manhattan Beach facility, also looks like Giant hardware and marker layout -feel free to correct me if I am wrong. If the unique poured concrete construction doesn’t give it away, they also released images with giant Avatar banners in the background.


posted by Chris at 1:34 AM  

Monday, July 7, 2014

Geodesic Voxel Binding in Maya 2015

If you’re like me, your ears will perk up at any technology promising a better initial skin bind. So I decided to take a look at the new geodesic voxel binding in Maya 2015, I couldn’t find much information about it online, so I decided to do the usual and write the post I would have wanted to find when I googled. I hope it’s useful.


This new way of skin binding was presented by Autodesk at SIGGRAPH 2013.


Here’s a link to the SIGGRAPH 2013 white paper: Geodesic Voxel Binding for Production Character Meshes, definitely worth checking out. I do like how Autodesk is now using the word ‘Production’ a lot more. It seems they are no longer using simple test cases to test pipelines and workflows. Above, they used our Nanosuit, from the Crysis franchise. Here’s the full video that accompanies the talk: [LINK]

How It Works


The basic idea is that it voxelizes characters into three types of voxels, skeleton, interior, and boundary. This way it tries to eliminate cross-talk. At ILM we had a binding solution in Zeno that used mesh normals and this eliminated crosstalk between manifold parts like fingers, but most of this paper focuses on skinning non-manifold meshes, meshes with intersecting parts, open holes, etc.

In Practice

Here’s the hand of the Marius bust we send out for rigging tests, notice when skinned with Closest in Hierarchy, there is some significant crosstalk:


Here’s an initial finger bind with the new algorithm, there’s still some crosstalk at 1024voxel resolution (highest possible), but it’s much better:


As someone who is very nitpicky about my skinning, *any* crosstalk at all is unacceptable, and it takes me about the same amount of time to clean the tiniest values as it does these larger ones. Here’s a closer look at some of the crosstalk from the ‘gb’ binding:


Crosstalk isn’t just bad for deformation, but these tiny little values are inefficient and sloppy, especially if you are sending it to a game engine.

Another area that requires significant cleanup is the underarm area where the serratus anterior lies, here I thought the new approach would work very well, unfortunately the binding didn’t have a noticeable difference from previous methods.

click to enlarge

click to enlarge (Head mesh from CryENGINE Asset Pack on Steam)

Few things are more difficult to skin than the human face. Here you can see traditional vs geodesic. I will say it’s definitely better than the old bind, but still has issues. This is one of the first initial skin binds on a closed-mouth neutral bindpose I have seen that has no cross-talk on one lip. I tweaked the falloffs doing three different binds on the traditional on the left.



Another thing I like is a hint at a multi-threaded future. The binding process (voxel calculation, etc) is multi-threaded. At Crytek, we even make hardware purchasing decisions based on Maya not being multi-threaded. We get animators the fastest 2 core CPUs, this allows them better interactive framerates, and still a second core for a headless mayapy to export a long linear cutscene or animation. It’s nice to see Autodesk begin to think about multi-threading tools and processes.

In Conclusion

The new Geodesic Bind algorithm from Autodesk is a step forward. There’s still no free lunch, but I will be using this as my default bind in the future. I will update this post if I run into any problems or benefits not outlined here. It would be great if there was a voxel debug view, or the ability to dynamically drive voxel resolution with an input like vertex colors a map, or polygon density.

Backwards Compatibility: New Nodes and Attrs

If you just want to use the latest Maya to try the feature, here are some gotchas. There is a new geomBind node, and some attributes on shape nodes:

// Error: file: C:/Users/chris/Desktop/TechAnimationTest/TechAnimationTest/ line 28725: The skinCluster ‘skinCluster1’ has no ‘gb’ attribute. //
// Warning: file: C:/Users/chris/Desktop/crytek_sdk_head_a/ line 27464: Unrecognized node type ‘geomBind’; preserving node information during this session. //
// Error: file: C:/Users/chris/Desktop/crytek_sdk_head_a/ line 34: The mesh ‘eyes_MSH’ has no ‘.sdt’ attribute. //

The geomBind node stores ‘the post voxel validation state performed during the geodesic voxel bind algorithm.’ and some other attributes. It has a message attr that connects to a skinCluster. The SDT attr on shapes is not related to skinning, it is a new ‘Subdivision Method’ attr for the openSubDiv support.


The above said, it seems to work fine for me if I just delete that stuff, the skin weights are fine.


posted by Chris at 1:29 AM  

Tuesday, May 27, 2014

PyQt: Composite Widgets


So the past few nights I was racking my brain a bit to get multiple widgets adding to a listview. I wanted to see a list of animations, each item in the list needed to have clickable buttons, and special labels.

I scoured the internets, and dusted off my old trusty ‘Rapid GUI Programming with Python and QT‘ book, I got the idea for the above from the ‘Composite Widgets’ chapter subsection, though they don’t use setItemWidget to insert a composite widget into another widget.

Here is what my QtDesigner file looked like:


I wanted to dynamically load a UI file of a custom widget and compile it with the UIC module. I first looked at making a delegate, but I just could not get that working, if you have done this with a delegate, let me know in the comments! (From the docs, it seems delegates cannot be composites of multiple widgets)

In the end I used pyuic4 to compile the above UI file into a python code, I dumped this, minus the form/window code, into a class I derive from QWidget:

class animItemWidget(QtGui.QWidget):
    def __init__(self, parent=None):
        super(animItemWidget, self).__init__()
        self.horizontalLayout_4 = QtGui.QHBoxLayout(self)
        #blah, blah, blah

At the bottom of that length UI frenzy of an init, let’s connect a button to a function:

self.connect(self.button02, QtCore.SIGNAL("clicked()"), self.awesome)

Now define that function, let’s just print that the animation that the widget in the list whose button you clicked is AWESOME:

    def awesome(self):
        print self.label.text() + ' is awesome!'

This could do anything with the anim name or various data bound to this object, like check out/sync a file from Perforce, load a file in Maya, etc.

Now let’s make our main window. We are going to use setItemWidget to insert our animItemWidget into the QListWidget called ‘list’. Notice that I have access to every UI element in the composite widget.

from PyQt4 import QtGui, QtCore, uic
class uiControlTest(QtGui.QMainWindow):
    def __init__(self):
        super(uiControlTest, self).__init__()
        self.ui = uic.loadUi('uiControlTest.ui')
        for i in range(0,100):
            wid = animItemWidget()
            wid.label_2.setText('Last edited by chrise @2014.06.21:23:17')
            wid.label.setText('Animation ' + str(i))
            wid2 = QtGui.QListWidgetItem()
            wid2.setSizeHint(QtCore.QSize(100, 40))
            self.ui.list.setItemWidget(wid2, wid)

Now, of course, in my example I just quickly made a bunch of widgets, so their names are all default, but you get the idea. If you have a better way to do this, perhaps something more performant, please let me know in the comments.

Note: It looks like that book is freely available on a college class website, save yourself 50 bucks:

posted by Chris at 2:38 AM  

Sunday, May 11, 2014

Maya: Vector Math by Example


This post is about how to use vector math and trigonometric functions in Maya, it is not a linear algebra or vector math course, it should give you what you need to follow along in Maya while you learn with online materials. Kahn Academy is a great online learning resource for math, and Mathematics for Computer Graphics, and Linear Algebra and its Applications are very good books. Gilbert Strang, the Author of Linear Algebra, has his entire MIT Linear Algebra course lectures here in video form. Also, Volume 2 of Complete Maya Programming has some vector math examples in MEL and C++.



Think of the white vector above as a movement. It does have three scalar values (ax, ay, az), sure, but do not think of a vector as a point or a position. When you see a vector, I believe it helps to imagine it as a movement from 0,0,0 – an origin. We don’t know where it started, we only know the movement.

A vector has been normalized, or is considered a unit vector, when it’s length is one. This is achieved by dividing each component by the length.


There are many Python libraries dedicated to vector math, but none ship with Python itself. I have tried numPy, then pyEuclid, and finally piMath. It can definitely be a benefit to load the same vector class across multiple apps like Maya, MotionBuilder, etc.. But, I used those in a time when MotionBuilder had no vector class, and before Maya had the API. Today, I use the vector class built into the Maya Python API (2.0), which wraps the underlying Maya C++ code: MVector

I had to call out 2.0 above, as those of you using the old API, you have to ‘cast’ your vectors to/from, meaning that classes like MVector (Maya’s vector class) don’t accept python objects like lists or tuples, this is still the case with the 2014 SWIG implementation of the default API, but not API 2.0. One solution is to override the MVector class in a way that it accepts a Python lists and tuples, essentially automatically casting it for you:

class MVector(om.MVector):
    def __init__(self, *args):
        if( issubclass, args, list ) and len(args[0])== 3:
            om.MVector.__init__(self, args[0][0], args[0][1], args[0][2])
            om.MVector.__init__(self, args)

But that aside, just use Maya Python API 2.0:

#import API 2.0
import maya.api.OpenMaya as om
#import old API
import maya.OpenMaya as om_old



Let’s first create two cubes, and move them

import maya.cmds as cmds
import maya.api.OpenMaya as om
cube1, cube2 = cmds.polyCube()[0], cmds.polyCube()[0]
cmds.xform(cube2, t=(1,2,3))
cmds.xform(cube1, t=(3,5,2))

Let’s get the translation of each, and store those as MVectors

t1, t2 = cmds.xform(cube1, t=1, q=1), cmds.xform(cube2, t=1, q=1)
print t1,t2
v1, v2 = om.MVector(t1), om.MVector(t2)
print v1, v2

This will return the translation in the form [x, y, z], and also the MVector, which will print: (x, y, z), and in the old API: <__main__.MVector; proxy of <Swig Object of type ‘MVector *’ at 0x000000002941D2D0> >. This is a SWIG wrapped C++ object, API 2.0 prints the vector.

Note: I just told you to think of vectors as a movement, and not as a position, and the first example I give stores translation in a vector. Maybe not the best, but remember this translation, is really being stored as a movement in space from an origin.

So let’s start doing stuff and things.


We have two translations, both stored as vectors, let’s get the distance between them, to do this, we want to make a new vector that describes a ray from one location to the other and then find it’s length, or magnitude. To do this we subtract each component of v1 from v2:

v = v2-v1
print v

This results in ‘-2.0 -3.0 1.0’.

To get the length of the vector we actually get the square root of the sum of x,y,and z squared sqrt(x^2+y^2+z^2), but as we haven’t covered the math module yet, let’s just ask the MVector for the ‘length’:

print om.MVector(v2-v1).length()

This will return 3.74165738677, which, if you snap a measure tool on the cubes, you can verify:


Use Case: Distance Check

As every joint in a hierarchy is in it’s parent space, a joint’s ‘magnitude’ is it’s length. Let’s create a lot of joints, then select them by joint length.

import maya.cmds as cmds
import random as r
import maya.api.OpenMaya as om
root = cmds.joint()
jnts = []
for i in range(0, 2000):
    jnt = cmds.joint()
    trans = (r.randrange(-100,100), r.randrange(-100,100), r.randrange(-100,100))
    cmds.xform(jnt, t=trans)
cmds.parent(jnts, root)


So we’ve created this cloud of joints, but let’s just select those joints with a joint length of less than 50.

sel = []
for jnt in jnts:
    v = om.MVector(cmds.xform(jnt, t=1, q=1))
    if v.length() < 50: sel.append(jnt)



The dot product is a scalar value obtained by performing a specific operation on two vector components. This doesn’t make much sense, so I will tell you that the dot product is extremely useful in finding the angle between two vectors, or checking which general direction something is pointing.

dot = v1*v2
print dot

USE CASE: Direction Test


The dot product of two normalized vectors will always be between -1.0 and 1.0, if the dot product is greater than zero, the vectors are pointing in the same general direction, zero means they are perpendicular, less than zero means opposite directions. So let’s loop through our joints and select those that are facing the x direction:

sel = []
for jnt in jnts:
    v = om.MVector(cmds.xform(jnt, t=1, q=1)).normal()
    dot = v*om.MVector([1,0,0])
    if dot > 0.7: sel.append(jnt)

USE CASE: Test World Colinearity

This one comes from last week in the office, one of my guys wanted to know how to check which way in the world something was facing. I believe it was to derive some information from arbitrary skeletons. This builds on the above by getting each vector of a node in world space.

def getLocalVecToWorldSpace(node, vec=om.MVector.kXaxisVector):
    matrix = om.MGlobal.getSelectionListByName(node).getDagPath(0).inclusiveMatrix()
    vec = (vec * matrix).normal()
    return vec
def axisVectorColinearity(node, vec):
    vec = om.MVector(vec)
    x = getLocalVecToWorldSpace(node, vec=om.MVector.kXaxisVector)
    y = getLocalVecToWorldSpace(node, vec=om.MVector.kYaxisVector)
    z = getLocalVecToWorldSpace(node, vec=om.MVector.kZaxisVector)
    #return the dot products
    return {'x': vec*x, 'y':vec*y, 'z':vec*z}
jnt = cmds.joint()
print axisVectorColinearity(jnt, [0,0,1])

You can rotate the joint around and you will see which axis is most closely pointing to the world space vector you have given as an input.

USE CASE: Angle Between Vectors


When working with unit vectors, we can get the arc cosine of a dot product to derive the angle between the two vectors, but this requires trigonometric functions, which are not available in our vector class, for this we must import the math module. Scratching the code above, let’s find the angle between two joints:

import maya.cmds as cmds
import maya.api.OpenMaya as om
import math
jnt1 = cmds.joint()
jnt2 = cmds.joint()
cmds.xform(jnt2, t=(0,0,10))
cmds.xform(jnt1, t=(10,0,0))
root = cmds.joint()
cmds.parent([jnt1, jnt2], root)
v1 = om.MVector(cmds.xform(jnt1, t=1, q=1)).normal()
v2 = om.MVector(cmds.xform(jnt2, t=1, q=1)).normal()
dot = v1*v2
print dot
print math.acos(dot)
print math.acos(dot) * 180 / math.pi

So at the end here, the arc Cosine of the dot product returns the angle in radians (1.57079632679), which we convert to degrees by multiplying it by 180 and dividing by pi (90.0). To check your work, there is no angle tool in Maya, but you can create a circle shape and set the sweep degrees to your result.

Now that you know how to convert radians to an angle, if you store the result of the above in an MAngle class, you can ask for it however you like:

print om.MAngle(math.acos(dot)).asDegrees()

Now that you know how to do this, there is an even easier, using the angle function of the MVector class, you can ask it the angle given a second vector:

print v1.angle(v2)

There are also useful attributes v1.rotateBy(r,r,r) for an offset and v1.rotateTo(v2). I say (r,r,r) in my example, but the rotateBy attr takes angles or radians.

CHALLENGE: Can you write your own rad_to_deg and deg_to_rad utility methods?

USE CASE: Orient-Driver

Moving along, let’s apply these concepts to something more interesting. Let’s drive a blendshape based on orientation of a joint. Since the dot product is a scalar value, we can pipe this right into a blendshape, when the dot product is 1.0, we know that the orientations match, when it’s 0, we know they are perpendicular.


We will use a locator constrained to the child to help in deriving a vector. The fourByFourMatrix stores the original position of the locator. I tried using the holdMatrix node, which should store a cached version of the original locator matrix, but it kept updating. (?) We use the vectorProduct node in ‘dot product’ mode to get the dot product of the original vector and the current vector of the joint. We then pipe this value into the weight of the blendshape node.

Now, this simple example doesn’t take twist into account, and we aren’t setting a falloff or cone, the falloff will be 1.0 when the vectors align and the blendshape is on 100% and 0.0, when they’re perpendicular and the blendshape will be on 0%. I also don’t clamp the dot product, so the blendshape input can go to -1.


The cross product results in a vector that is perpendicular to two vectors. Generally you would do (v1.y*v2.z-v1.z*v2.y, v1.z*v2.x-v1.x*v2.z, v1.x*v2.y-v1.y*v2.x), ut luckily, the vector class manages this for us by using the ‘^’ symbol:

cross = v1^v2
print cross

USE CASE: Building a coordinate frame


If we get the cross product of v1^v2 above, and use this vector to now cross (v1 x v2)x v1, we will now have a third perpendicular vector to build a coordinate system or ‘orthonormal basis’. A useful example would be aligning a node to a nurbs curve using the pointOnCurveInfo node.


In the example above, we are using two cross products to build a matrix from the tangent of the pointOnCurveInfo and it’s position, then decomposing this matrix to set the orientation and position of a locator.

Many people put content like this behind a paywall.
If you found this useful, please consider buying me a beer.

posted by Chris at 11:49 PM  
Next Page »

Powered by WordPress