View Full Version : which is the most under developed rigging tool?

05 May 2008, 05:55 AM
I am an animation student and I am doing some research on rigging. I am wondering which you think (relatively) is the most under developed rigging tool in Maya?

Looking at all kinds of plugings now a days I think they complemnt a lot with Maya's built in tool. I know that muscle is one topic that is addressed a lot in many plugins.

05 May 2008, 09:23 AM
Everything that helps mirroring stuff. whether it is geometry, joints, poses, animation etc. Not only while creating, but also AFTER they have been created. So for example you have rigged a character and later the animators say: Oh this joint here needs to be more like this so we get a better arc/deformation whatever. How will you be able to edit (symmetrically) your hierarchy without messing around forever?

05 May 2008, 11:08 PM
That's interesting. I have had some problems with mirroring esp with the joint mirroring and geo mirroring. But nothing very serious so far. In your example do you want to change things on both the sides (i.e. refect the changes on other side)? or You want to just move a skinned joint?

05 May 2008, 05:51 AM
The bones,. they highlight everything under it.. (not intuitive)
Adding or removing influences,... (weighting disperses across multiple vertices) {locking/unlocking?.. wtf?}
CopyWeights, .. should work on any frame, not at zero,.. (am i the only one here to notice this?)
A bounding box "setup machine" style of locating mesh boundaries for the mesh deformation.
Automation of clavicles, and twist bones?.. Because animators never have enough time for that on most production cycles.
Easier import/export on anim or fbx for animation and new add of bip files cvs files, or bvh files for mocap.

Buexe?? Didn't you have something in a previous thread redoing Maya's api (is that you?.. thats awesome stuff.)
Maya needs that.


05 May 2008, 07:07 AM
I agree highlighting joints is annoying for me too.

For copying weight I am guessing that since your frame 0 is the bind pose, Maya will copy it. But any frame other than 0 (if you have animated) might not be the bind pose so maya's smartness doesnt help there.

Your clavicle suggestion is interesting too. I am noting it down.

Yes that's Buexe, but the video link is not working. could somebody post that video? I am very curious to see it after reading that thread. Seems like Buexe has done some cool stuff.

Thanks for your comment.

05 May 2008, 08:36 AM
Sorry, for the video being down, I`m trying to get a CMS on my site and deleted most "old" stuff on it. I don`t have a copy lying around here and the backup is in another country, if somebody has a copy I could upload it to another place. But generally it is not a huge thing that I have done there. I just felt that this whole auto-rigger-MEL-script-thing to be well ... a little bit sub-optimal. I have written two complete auto-riggers in MEL and while I certainly have learnt a lot by doing so, I always found it annoying how much effort it costs if you want to change stuff on the character`s proportions for example. Another thing is this space-switching thing. While you can see a lot of nice rigging reels that have some sort of space-switching, usually it is predefined and limited. FK/IK Blending switching is an art in itself and another "feature" that is usually predefined: an IK chain will run from the shoulder to the wrist, if it is a good rig you have some sort of matching/blending solution with a nice marking menu. So I thought if there could be a way to create a rigging system for different character types ( bipedal, quadrupedal + all sorts of variations ) that does not limit the animator to the predefined stuff ( for example run the IK Chain from clavicle to elbow and later from shoulder to wrist in the same scene) and gives TDs time to actually work on the more interesting stuff like muscles, advanced deformations etc.. This system is now in a stage where it can actually be used create, edit and animate all sorts of rigs. Hopefully, there will be a public version someday, the folks on cgtalk will be the first ones to know : )

@refract: what do you mean by "Automation of clavicles"?
And reading refracts comments I`d probably have to say now a good, solid, "multiple references crazy names and many characters in a scene" proof animation im- and exporter is probably something many people wish for.

05 May 2008, 06:45 PM
No problem buexe. But I would love to see what you have done, so anytime if you get hands on those files it would be great if you can post it. And again thanks for sharing.

I agree with you on changing proportions. I dont have much rigging experience but so far I have found that there are many things in rigging which after one step or the other very difficult to modify. And you always have to be mistake proof from the beginning or else you have to just redo a lot of stuff if you find a mistake later.

In your example of IK running from clavicle to elbow I didnt get the exact purpose. You mean the same joint could be involved in multiple IK solvers? What is your purpose of doing so?

And I totally agree with space switching. I am always impressed with that feature, but what I later found and made me not so impressed by it is that they are usually fixed, having blend between constraints.

I hate that importing references with long names.

05 May 2008, 06:58 PM
Painting weights in Maya is atrocious. There's a plethora of 3rd party tools to try and make it bearable.

05 May 2008, 10:09 PM
In your example of IK running from clavicle to elbow I didnt get the exact purpose. You mean the same joint could be involved in multiple IK solvers? What is your purpose of doing so?
Say you have a character sitting at a table with it `s elbows resting on the table. Then your normal IK chain ( shoulder to wrist ) is no good and switching to FK is no good either. There are some folks using some sort of elbow locking feature (for example Jason Schleifer in his Animator friendly rigging DVD series) and as an alternative I thought it would be cool to have the shoulder act as an IK effector and the elbow would be the IK Handle. I don`t know how often such a feature would be really used, who knows. I just love the idea of having the option to really change the structure of a rig without having to write a script and/or fiddle for hours in the hypergraph.

05 May 2008, 11:12 PM
Cool. Thanks Buexe, I got what you are trying to say. I tried placing two IK chains in maya similar to what you said. I did two chains, clavicle to elobow and shoulder to wrist. It looks weired the way it's working so far or probably by some proper heirarchy we can make it work. I think you must already have tried by using existing maya IK. So you cant achieve that elbow locking using current IK and setting up the way you mentioned. What do you think is the main problem? (sorry if I am asking too many questions, but I am just starting up and I feel like I am millions years behind in knowing rigging stuff)

05 May 2008, 11:13 PM
Hey Ben, Sometiems it annoys me, but so far I am not too disappointed with painting weights. I didnt know there were plugins which do better jobs for this task. Do you have any particual plugin in mind you can suggest that I can try? Thanks.

05 May 2008, 12:47 AM
As far as I know there aren't any plug-ins specifically for painting weights, but there are a bunch on highend that list the joints in a more convenient (larger) window, allow one click holding and unholding of multiple joints, selecting verts based on influence, copy and paste/paste average weights, etc.

My issues are mainly with how the skinCluster node deals with normalizing. It has a tendancy to add random weights to odd joints, which necessitates the holding etc. The component spreadsheet is kind of clunky, and so on. IMO XSI has the best weight painting interface, although I've never tried Lightwave or Houdini.

05 May 2008, 01:23 AM
That's quiet intereting ideas for painting the weights. And I agree with you on normalizing parts. I had a very frustrating time, skin would add vertices of head under spine influences or anything randomly.

XSI apparently seems to be a very strong competitor in rigging and animation.

05 May 2008, 07:16 AM
Hey Beuxe,. CAT,. character animation technologies,.. is somewhat like character studio,. I liked your setup because it was so close to this. Softimage bought CAT. Not sure where it is now. I exclusively use Maya now,.. and Maya really needs a new bone setup. Your setup looked really close to CAT and character studio for Max. Very cool shit.
ANY alternative bone setup for maya would be amazing. And the ability to load bip files or fbx, or .anim files for it.
(I just want to see another update on your work.. I think ur in the right direction)////....


05 May 2008, 03:29 PM
does anybody have any experience with animation master?

I was watching sample video of jef lew's tutorial and It seems so easy to move the joints in animation master. Here is the link to the video
(I found the tut sample really great, gonna buy those DVDs)

Now I am very eager to see what Buexe has done.

05 May 2008, 09:47 PM
For smooth skin editing I always use dwSkinningTools ( David Walden is the author ) because there are features to quickly edit the "hold" on many/all influences at once. Setting the hold is crucial to edit smooth skin weights in a rational predictable manner IMHO.

Regarding CAT I have looked at it in the past and must say that I liked it a lot. I give it a quick spin once and liked the modular approach. Seems overall a very powerful package. Motionbuilder and Character Studio also have great features but I always felt that their system is constraining too much to a certain type. Like you can add extra stuff to the HumanIK thing but in MB but it will always be just an extra. It will not get the same behavious as if it was "inside" the humanIK solver, you know this dragging around thing. But I guess that`s always the problem with rigging systems. If you want to have a certain feature it costs flexibility on some other level. In this experimental thing I developed there is a high degree of flexibility (bi-directional constraining, even invert complete hierarchies, IK chains where you want them, character>character interaction ). But this comes at the cost of giving the animator responsibility to act careful and not expect that the system will do anything. But it is really nice to hear folks are actually interested in this, I`m working on it. But please understand that I put myself under an NDA regarding demo stuff until I have more than just an experimental monster eating my spare-time. But I`ll put on the old vid if I can get hold of it, although I really think it is not such a big deal.

05 May 2008, 10:19 PM
Buexe I appreciate your sharing this with us. And I should thank you along with other folks for giving me motivation to think more.

Since you have explored a lot of stuff already and have more experience you might not feel it as a big deal, but for people who are new and not much experience will always find stuff that you are doing amazing and inspiring.

05 May 2008, 10:47 PM
thanks, theflash : ) but if you allow me to give you an advice ( not that I think you need one though). Check out the Jason Schleifer Animator friendly rigging bundle. It is really well done and will give a lot of nice info to create a more than decent 3d character for animation. I wish I had a source like that when I started, then I would have probably skipped all this MEL API mumbo-jumbo.

05 May 2008, 04:41 AM
Thanks for the advise. I have heard a lot about his DVDs and your advise for referring them adds to my desire to check the DVDs. I love his stuff, and I want to be like him who knows rigging and loves to animate characters. Probably his love for character animation drives his rigging skills further.

I think over the course by experimenting you must have learned a lot from inside out of maya. I would love to get that experience :)

What I am planning for summer is to do some work on rigging tools that needs more attention. For muscles there are some really cool stuff already out there (Mr Comet's is the best so far). So I am searching for some new topics.

CGTalk Moderation
05 May 2008, 04:41 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.