Reusing rigging information for other characters?


Well, ideally your high res rig should deform perfectly all the time so you don’t need to check it during animation :slight_smile: The practice is however that things go wrong, but you should fix the rig and not the animation in this cases…
Anyway, we’re usually working with mocap here, so there’s no option to get the mocap talent to check if he’s creating good deformations - it just has to work out somehow in the end :slight_smile:



actually, I like to use the referencing script that I wrote to just reference in different animation resolutions… and then I have a totally separate skinning file which takes the animation and does all the rendering with that. By having different resolution skinning models I can ensure that I’ve got a version of the rig for just blocking, one for adding more detail, and one for finalizing. That way I know I’m not getting ahead of myself when animating, as I can’t animate stuff that I shouldn’t animate yet.



Our character rigs contain both the skinned “high rez” version as well as the segmented-and-parented-to-bones “proxy” version of the character. We can toggle between the proxy or high rez with a switch, and we have separate switches for the head and body. Since the high rez version just has blendshapes and smooth skinning usually, it isn’t all that slow to turn on or too heavy on the scene. All our character rigs are referenced into the scenes so that the scenes themselves don’t become all that large. The animators will usually animate in proxy mode, and then switch to high rez if they need to check the deformation closer and for rendering previews. When the high rez geometry is hidden, Maya naturally turns off the deformers so we don’t have to manually turn them off ourselves. Since we aren’t doing cloth, muscle, skin, or hair simulations we don’t have need for a slower “simulation” rig.

For my own personal rigs and for future rigs that might require simulation, I’m going to be exploring an approach similar to what Jason talked about in his Fast Character Rigs DVD (I was at his SIGGraph talk the DVD is based on, but don’t have that DVD itself.) Actually I’m going to be exploring an approach where the deform rig (parenting, skinning, or simulation) is separate from the control rig (which holds all the animation). Then the deform rig would be loaded and constrained to control rig, or switched out with another deform rig as needed, while the control rig would always remain in the scene. Dunno how well it will work or if the constrain nodes will slow it down too much and outweigh the benefits.

Michael Duffy


Jason: Aah, thats what the script did, I was close:rolleyes:

Dont you wanna give some tips on how you skin characters. Thats probably the number one problem Im having and there is’nt really alot of resources out there for skinning.



Someone hit the nail on the head earlier…
This thread is PURE GOLD! :slight_smile:
Can’t thank you enough for your generosity and valuable time, Jason…
You’re one helluva Guy! :slight_smile:



Most animators have a really hard time building re-useable character rigs allowing them to pre-built animation clips.

This is why at Kaydara we have built a re-useable character rig that works with any character sizes and proportions and has automatic features such as hand/feet floor contact.

MOTIONBUILDER works with any 3D app (max, LW, maya, xsi) and is available until July 31st at an amazing promotion of $100 for a one-year license (we will offer to all those who purchased this license to convert to a lifetime permanent license for another $100)




I feel like a newborn fish with a pituitary disorder, who was just caught out of the creek by aliens and taken to a vast planet of purple ocean and living silicon based rock lifeforms…

(Ritchie hugs his lightrigs)

Jason, I was wondering. I asked you a while back and I think you touched on it for me… but I have this huge honking beast based from a lovecraft nightmare that has like a whole host of tentacles and weird slimy nodules. It will take me a while to model since I am more of a set modeller, lighting td and rendering kinda guy… but was just wondering, is there an easier way to rig tentacles other than using a ton of joints and doing the stop motion animation joint by joint keying trick? I mean this thing is going to have so many tentacles my computer might die. So to relate this to the topic (lol) is it at all possible to reuse joint rigs within the same character model?

Maybe I should stick to sets and lighting. :blush:

Ritchie Roberts
ZeroNeuro Arts Ltd.


you can rig 'em with wire deformers, which might help… or use a bunch of influence objects that the animator just moves around… or you can use cloth/syflex… or spline ik… it all depends on the type of control you want the animators to have.

For the watcher inthe water (lord of the rings, film 1) we used a combo of spline IK and fk depending on what the animator wanted. it worked really well! :slight_smile:



Well I see this tread is changing towards skinning, wich is a great thing. And speaking of skinning. What would be the most productive and ifficent way of skiining a verry detailed, smoth poly model?

I mean do you attach the smoth rez directly to the joints, or do you have, like a unsmoth version of the smoth rez and skin that to the joints and add all the muscle work and influence object to that perticular unsmoth rez. And then attach that to the smoth version?

If: Yes. What method would you recomend?

If: No. Y not?


Hello people! :cool:

Jason, u were saying on the “Integrating A Creature Whithin A Production Pipeline” dvd something about a script that keeps the the spline ik curve the same length. Could you tell me where i can find that script or something similar or even a cheap tip will be cool .




Hey Gabriel,

I’d suggest going to Jason’s website and asking this question on his forum. It’s got a section just for questions on his DVD. Here is the link:

Jonh and his dog



Originally posted by bentllama
Most of my rigging for bipedal creatures is automatic. The beatuy of MEL is that you can script the usual grunt labour you repeat all the time…

Do you have to adjust the weights for the joints greatly? Or is the automatic weighting near perfect?

In Blender we have a new auto skinning feature. I have yet to use it. Some folks say it works perfectly but it’s a new feature so I have not made use of it. The auto skinning process seems to be the way to go if it works ok in an app. Manual is always more reliable. In 3d reliable methods are often slow.


From what I’ve experienced the procedural stuff never quite does everything for you to prefection, but it does do a vast majority of work that otherwise would become tedious and boring. There is always a bit of manual tweaking and stuff you need to do in order to get really good results. Anybody that is looking for a one button perfect rig is asking for a bit too much :slight_smile:

Procedural muscle rigs are a really good starting point to much better deformation though.



hey Jason,

thx for all the postings, even though I use max.

My main problem is getting some of the vertices on the fingers and toes unstuck. I’m guessing this is an amateurs problem, probably nothing any of you get phased by.

cheers man!


Originally posted by nobrain
[B]hey Jason,

My main problem is getting some of the vertices on the fingers and toes unstuck. I’m guessing this is an amateurs problem, probably nothing any of you get phased by.

cheers man! [/B]

I think that it is always best to work in stages and save often. I have ruined many a rig while trying to setup a working rig. Most animators use some method of assigning weight groups often before a rig is applied. At that time you can make sure that every part is assigned. You can use automatic weighting if your app allows this. It is so much easier to adjust weights later on.


Speaking of skinning :slight_smile:

On a “naked” character biped (good examples are Cave troll and Gollum), how were seem problem’s overcome between the body and head?

Is it practical to have one single Sub-D mesh (head and body and arms/hand, legs/feet all one mesh) over your rig/deformers ?..or is there clever ways to eliminate seem issues in SubD’s.

I guess I worry about having a very large and complex subD mesh, some areas being effected by influence objects or as a Wrap (body/arms/legs) , and some areas effected by Clusters and Blendshapes (Face, neck, hands)…however, is this how it is “done” on the whole?

Be great to know, as It’ll point me in the right direction to find a solution for my own Character issue :slight_smile:

As it is a Naked character, there is no segmenting or clever disguising of seems with bits of clothing/collars etc.

Cheers for any thoughts anyone might have on the topic :slight_smile: (and yes, think it is related to the topic of Rigs :wink: )


Max does support this actualy and does a great job of it. You can do it in several ways.

if you are using the same rig with similar naming conventions you can save the envelope and weight data out to and file and load it back into the new character. If the vertex count and order has changed and you have weighted the verts with paint weights or weight table then this method will not work.

Another method if every thing is the same is to just copy the Skin modifier off of one character and paste it on another. We do this all the time to update a character mesh in an already animated scene.

Next method is more flexable because the vert count, size and shape of character can change all that you like. In the Tools Panel go to more and open SkinUtilities. Here you can select the original mesh, press bake to mesh, which will create a mesh with all the saved Skin data stored in it. Merge this mesh with the new character. Add Skin to the new character and add the bones to it. Select both the new mesh and the baked mesh and in the SkinUtilities press Import Skin Data From Mesh. Just so you know you can remodel the skin data mesh any way that you want to make it fit the new mesh. the closer that it is the better the result will be. You can also use this tool to extract weights from one side of a character and paste it to the other. You would only need to do this if the character wasn’t even close to being a mirror copy. You can just model up the data mesh to fit the other side and then import the data from it.


I wanted to say thanks for all the information Jason. I do allot of rigging in both Max and Maya and your process is great to hear. I do most of my work in the TV series market and some feature work.

As for the discussion being maya focused that is fine. Max and maya as well as others are very similar in what you can do but getting there is often quite different. If you know other packages you can usualy apply this information with a few changes to reflect different tools.

Once again thanks, and thanks for all your great work.


Another method for dealing with high and low res rigs like what Jason was talking about is to use constraints instead of copying animation.

So the animators work with the animation rig that is fast and doesn’t require all the extras that you need for skinning a character. The TD’s work with an FK rig that has all the extras and they skin to that rig. Merge the skin rig into the animation file and constrain the skin rig to the animation rig. This allows the animator to see both at once and even make small changes to the animation while seeing the final character. This isn’t something that you would want to do with a character like gullom as he is so complex but I work mostly on TV series in Max and this is a very fast workflow since we can merge and delete the final character at any point in the animation process. I have a full set of Max scripts that run this and make it seamless to the animators and lighters.

A host of scripts for speeding up rigging is also the answer to all your rigging problems.


does anyone try to use Houdini for character? I heard that the digital assets does the same thing that you guyz want.