Rigging

[Gerry]

I've been working through the Barry Zundel videos but keep running into a wall. I've learned a ton, but I think my time will be better spent getting back to the "Nightcallers" trailer.

 

I think I sort of understand rigging better than I did before, but in the videos he encounters occasional problems (mostly with Compensate Mode and parenting hierarchy) and it just gets my mental knickers in a twist.

 

I don't need the rigging knowledge for "Nightcallers" since I use The Setup Machine, and I am making good progress on modeling Biff, so it's back to them lovable bugheads for the nonce!

[robcat2075]

Can you describe the problem?

[Gerry]

It has to to with the arms and the control rig mainly. There are issues with "Compensate Mode" in the version he's teaching in, and as much as I tried to follow meticulously, there are a few points where having it "off" by default causes him to go back over and fix things, reparenting, etc, when, e.g., "attached to parent" doesn't properly show in the properties panel.

 

I also deviated at a couple of points that didn't seem *important* at the time (moving a couple of bones, or fiddling with CP assignments) after constraints were created, but neglecting to note that some constraints were "on" by default in the model instead of "not set". Seems like some of those decisions might have been poorly thought out!

 

He also takes some very efficient "workaround" routes, e.g., not using automatic CFA but instead copying over geometry, bones, constraint relationships, CP assignments, etc. manually, though in a highly organized way that I understood and could follow along. David was very helpful about fixing the arm, but my whole aim was to get it myself, not have others fix my mistakes.

 

But that's all details. I'm unable to really "see through" to the mechanical cause-and-effect and comprehend the big picture of what's going on. I could post my wonky model but then I'm back to my original question, which is, should I be taking the time for this or get back to my live project. This morning I decided to spend the day working through the rig from scratch but once I started on that path and started poking through the last few saved versions of the model, I had no idea where to start.

[robcat2075]

I hope to have a good explanation of "why we rig" in NewTAoA:M so that everyone can build a simple, functional rig from scratch and understand how it works and... learn enough from that so "installer" rigs make sense as the better option.

 

I really think "installer" rigs are better and faster for most uses, but people seem to get stopped dead by them for a variety of reasons.

[Gerry]

I'm not sure there's a "simple" solution. It reminds me of what's required for creating a blog. There are packages (I'm familiar with WordPress but I have a feeling they're all similar) and the basic problem is this: They try to create a blog formatting interface that requires no actual code writing; but what they end up with is something that is every bit as complicated as writing code, the only difference is there's no code writing. But it's not a simpler solution.

 

Rigging strikes me as similar. You can either tackle rigging, or you can tackle installing a pre-made rig. But either way there are many complex steps and no one-size-fits-all solution.

 

I've installed the Setup Machine a good half-dozen times, but every time I do, I open up the instructions, read through them more than once, and keep them open while I'm installing. I suppose if I did nothing but this it would come more naturally, but otherwise, it's always a complex exercise that requires intense focus.

[largento]

I'm totally with you on this one, Gerry. It's partly why I've opted for simpler characters. :-)

 

I found Barry's rigging tutorials the least helpful of the set. That's not saying that they didn't convey a great deal of information on the basics of bones, but it seemed clear that Barry was following his notes, too ...and as you pointed out, making lots of mistakes that make it frustrating to follow along. And then Barry sort of skips over assigning control points, which was one of the great mysteries for me early on.

 

Errol Flemm was rigged a few different times and it really put the brakes on my productivity. (I ended up rigging him with the Setup Machine the last time around.) Big thanks to Mark S. for his help and his rig, because I would probably still be stuck. :-)

 

Myself, I approach animation more like stop-motion, so I'm okay with just using FK for everything and you can create a very simple rig requiring no cosmetic or control bones if you're okay with not having any bells or whistles. (I simply use cp-weighting for the joints.) It's not very elegant, but it can get the job done. I also find myself considering the rigging when I design/model characters and look for things that can make it easier (wristbands so that the hands don't have to be attached and twist the arms, etc.)

[robcat2075]

Tell me more about what you don't understand. It might get covered in a tut eventually.

[Gerry]

I found myself wishing for a graphical diagram showing all the bones, their relationships, constraints, the geometry bones vs. the control bones, etc. Some way I could see in a picture where everything went and why. I know that's a tall order, and made complicated by the fact that rigs are as individual as their creators. But it wouldn't have to be rig-dependent, just a live diagram interface where maybe bones could be dragged around and reordered. I know that's the intent of the pws but unfortunately I don't find that graphical enough. A diagram-based system like this would enable you to turn sets of bones on and off, and maybe by mousing over you could see constraints and parent/child relationships.

 

Other apps have a diagram-based way of showing stuff like this. the node system in Maya, or the new materials editor in 3DS comes to mind, but Adobe Encore has something similar and I believe After Effects does too though I just discovered that by accident a couple weeks ago.

 

I should stop now. Like Confucius said, be careful what you wish for.

[robcat2075]

I've seen things like that and while it should be helpful I can't help but note that exceedingly few Maya users learn to rig even with that graphic aid. They don't seem to gain much more understanding of "why we rig" with it or without it. I think fewer Maya users get their own characters rigged than A:M users.

 

Those charts look like lots of information is being presented in a clear format and yet it's only clear to the advanced user. A line from one bone to another that says "Rotation constraint" (instead of our current text listings) would speed me up quite a bit, but the user who doesn't grasp the purpose of the constraint isn't helped much. I think very very few people could look at a chart like that and predict what the rig will do.

[Gerry]

I've never gotten advanced enough with Maya to know if that flowchart interface is used for rigging. I was just making a general comment about their usefulness in presenting interrelated elements in a graphical way. So no, I'm not making a comparison about rigging in Maya!

 

EDIT: I should also say that with what I learned from the Zundel tutes, I feel like I could now build a basic rig from scratch. I think it's the "following along" that gets me in trouble, not the underlying theory of rigging. The one area where I'm still not sure of myself would be building squash and stretch into the limbs. The rig he builds doesn't have it.

[NancyGormezano]

The one area where I'm still not sure of myself would be building squash and stretch into the limbs. The rig he builds doesn't have it.

 

In A:M, I have found that in most rigs that claim to NOT have S&S (eg 2001 rig), one can usually animate the scale of the appropriate geometry bone(s) manually (and maintain volume by not scaling equally all dimensions) to do a pseudo S&S.

 

I have also found that the generalized INSTALL solutions, tend to introduce a level of complexity and way toooo many bones, that obscure and complicate the simplicity of rigging in A:M. They are good if the resultant rig does all that you need, but "May the gawds have mercy on your filet o' sole" if you need to debug, tweak, or add any special functionality that the rig doesn't address.

 

Holmes has some nice basic rigging video tuts specifically designed to address simplicity and basic concepts.

 

Yes, rigging is mind-boggling-disruptive and harmful to brain cells. Seems like I'm always starting from scratch (in terms of recall) if I haven't done it in awhile.

[robcat2075]

A diagram-based system like this would enable you to turn sets of bones on and off,

 

Are you talking about visibility?

[Darkwing]

This is only a lightly related question, is there a way to delete groups of bones at a time? What I mean is delete a parent and all it's children in one keystroke?

[NancyGormezano]

This is only a lightly related question, is there a way to delete groups of bones at a time? What I mean is delete a parent and all it's children in one keystroke?

 

Yes - but takes more than 1 key stroke. Assuming you have imported your model:

 

1) create also new blank model

2) shift click on the desired parent bone in "your model" - the one that you want to delete the bones

3) drag & drop to blank model

 

parent bone plus it's children are gone from "your model" and are now in "blank model". Delete "blank model".

[robcat2075]

This is only a lightly related question, is there a way to delete groups of bones at a time? What I mean is delete a parent and all it's children in one keystroke?

 

Not in one keystroke.

 

but if you select the parent (1st click), hit asterisk (2nd click) that will expand its tree, then SHIFT-select the parent bone at the top of the tree (3rd click) and the last bone at the bottom of the tree(4th click) then hit Delete (1 keypress), that will delete them all.

 

I don't think it would be possible to truly identify what you wanted to delete in only one keystroke.

 

 

Nancy's method is better than mine.

[Darkwing]

Well, what I meant by one keystroke was in only hitting Delete once. For example, I import a hand model but for whatever reason, don't want the rigging that was associated with it. So I go to bones and there's the main "Hand" bone of which everything underneath is a child. Clicking and deleting Hand doesn't delete it's children. More importantly though, what I wanted to know this for was say some of the child bones are themselves parents to other child bones (as is the case in a hand). Expanding Hand in the PWS and then shift selecting all of those bones, then makes me have to do it to every child parent that was in Hand as well. I hope that clarifies it a little by what I meant by one keystroke, it's mainly in those cases where some of the children are also parents to other bones and I just want to get rid of them all at once, that sort of thing.

 

Also, I never would have thought to do something like Nancy suggested, I think I'll play with it a little next chance I get cause I find myself deleting bones fairly regularly...

[Gerry]

What Robert describes (expand the tree, shift-select, delete) is how Zundel does it in his videos, and it works fine. Yes it's more than one click but it's efficient and does what you're trying to do.

[Gerry]

A diagram-based system like this would enable you to turn sets of bones on and off,

 

Are you talking about visibility?

Yep, just so you can unclutter the diagram pretty much.

[mtpeak2]

Well, what I meant by one keystroke was in only hitting Delete once. For example, I import a hand model but for whatever reason, don't want the rigging that was associated with it. So I go to bones and there's the main "Hand" bone of which everything underneath is a child. Clicking and deleting Hand doesn't delete it's children. More importantly though, what I wanted to know this for was say some of the child bones are themselves parents to other child bones (as is the case in a hand). Expanding Hand in the PWS and then shift selecting all of those bones, then makes me have to do it to every child parent that was in Hand as well. I hope that clarifies it a little by what I meant by one keystroke, it's mainly in those cases where some of the children are also parents to other bones and I just want to get rid of them all at once, that sort of thing.

 

Also, I never would have thought to do something like Nancy suggested, I think I'll play with it a little next chance I get cause I find myself deleting bones fairly regularly...

 

Instead of importing the hand model, just copy and paste the geometry.

[robcat2075]

A diagram-based system like this would enable you to turn sets of bones on and off,

 

Are you talking about visibility?

Yep, just so you can unclutter the diagram pretty much.

 

I believe enhanced Selection filters work on models also so you could use those to show any set you want in the PWS, Shift-select them and un-vis them in one swoop.

[robcat2075]

For example, I import a hand model but for whatever reason, don't want the rigging that was associated with it.

 

Nancy's method has got to be the ultimate in quick and easy. One drag and they're gone.

[robcat2075]

Another observation... adding Squash and stretch to an individual bone in a rig... I would call that an advanced task.

 

I would be very surprised if a typical user of any other program knew how to do that. Certainly most people called "animators" would not have a clue.

[Gerry]

Here's a thought, what about a "show properties" box for bones similar to what we've got for the modeling phase. I noticed there's already an info box that shows in bones mode if you hit R, S or N, but currently it only shows the relevant scaling, translation or rotation numbers.

 

What about a box that shows:

Bone name

Parent

Child(ren)

Constraints

Etc.

 

The intent being a simpler way to see all the relevant info for a bone in one place.

 

Doable? Desirable? Practical? Redundant?

[robcat2075]

Here's a thought, what about a "show properties" box for bones similar to what we've got for the modeling phase. I noticed there's already an info box that shows in bones mode if you hit R, S or N, but currently it only shows the relevant scaling, translation or rotation numbers.

 

What about a box that shows:

Bone name

Parent

Child(ren)

Constraints

Etc.

 

The intent being a simpler way to see all the relevant info for a bone in one place.

 

Doable? Desirable? Practical? Redundant?

 

Doable, yes. Redundant... somewhat, in the sense that the info is already available somewhere, just not in one place.

 

I agree that having to keep an awareness of everything in your head is challenging (I find it boggling at times), but I don't think this display of info for one bone at a time will help much. I don't think it would advance novice riggers much.

 

Perhaps we could brainstorm on a display that would genuinely and elegantly give us a global view with meaningful information.

[robcat2075]

On TSM2, here's how I see it:

 

1- TSMBuilder puts geometry bones in your model. You scale them to fit one side of your model.

 

2a- TSMFlipper mirrors what you have from step 1 to the other

 

2b - YOU attach CPs to bones. Weight or fanbone or Smartskin as needed at the joints.

 

3- TSMRigger adds all the control bones, adds all the constraints and hides all the geometry bones and anything else you don't need to see to animate.

 

Character ready to go!

 

 

I think that's a straightforward process that doesn't require you to know what all the constraints are that make it work nor how to create them.

 

Why isn't it?

 

I really mean that... what part of that is not working for people?

[largento]

I really mean that... what part of that is not working for people?

 

On the Mac, none of it.

[NancyGormezano]

what part of that is not working for people?

 

I love, love, love the simplicity of installation for tsm, and love, love, love the concept of the modularity of adding parts

 

However - what I do NOT like is:

 

1) TSM is not officially supported in A:M. Can it be used in most recent versions? (ie ver 16 64 bit?). Have to have old version to do install, and use for modifying rig, and saving model. Eventually that will not work any more, as will lose features in model when going back to old version of A:M. Do not want to become dependant on something that will become obsolete, and is not supported. Does it work for MACs (ver 16)? I'm assuming it still works in ver 16-32 bit version? (haven't tried). Will perhaps eventually break when constraint mechanisms change in future versions of A:M. Do not like how it is flaky with relationships - ie error strings (was designed when things were just called poses)

 

Would like to see documentation of script language syntax, without having to hack, guess at. I had been sucessful once before at writing own trial scripts, but last time I tried (maybe ver 16?) couldn't figure out or debug why didn't work. I gave up.

 

2) Standard TSM rig installs a gawd awful # of bones. My major complaint of all generalized INSTALL methods (lite rig, 2008, Squetch all are guilty, some more so than others). Too many bones in there just to make installatiion process work. Too complicated to tweak once rig installed. Does not use Nulls for control. Do not like other controls it uses.

 

3) Gawd help you if you install a face rig (of own design) or some other part or modify it to use nulls and then need/want to remodel certain parts, and have to reinstall rig portion, because it's too complicated to modify, debug, understand after being installed.

[robcat2075]

If you're on the Mac TSM2 is not there for you, that's true.

 

TSM2 still works on the current 32-bit A:M. Since you can have both 32-bit and 64 bit installed, I see that as only a minor problem for PC users.

 

My question is directed to the PC users who have it, run it and are still stopped dead in their tracks by it.

 

That normal default rig that TSM2 installs should be more than enough for to handle any animation situation that someone who is starting out with their own characters will encounter. I know it's more than enough because I've done lots of animation with it.

 

My premise is that if you have made a character, you will be animating it sooner with TSM2 (or any "installer" rig) than if you try to learn rigging on your own and make one from scratch. Most beginners who try to build a rig from scratch will build a lousy one that is lousy to animate with and then complain that the program is the problem.

 

 

 

The scripting... it was never sold as something any user would tinker with on the script level so i think it's a cool extra that it can be done AT ALL. Changing the scripts is not something a beginner needs to do at all to get TSM2 to work or to use it.

 

Face rigs. You do have to do your own solution for face rigs with TSM because face rigs aren't part of it. But I've never had trouble putting my face rigs on it. The only bone up there is the head bone, navigating that and making your face bones children of that shouldn't be a big issue.

[Gerry]

As we speak I'm installing TSM2 and it works fine, install is easy and customization is easy. I have no real complaints with it although it doesn't currently work past v.15.

My characters are completely non-standard and I still puzzle over exactly how to rig shoulders and hips though it's way simpler than it would be for a more humanoid character. Likewise faces are even more non-standard than bodies.

 

I have no complaints with TSM but I would like to comprehend rigging more thoroughly, and the Zundel tutes have done that even though the model I built and rigged is pretty wonky. that's not my concern. It's no different than having to do a whole bunch of crappy drawings to get to one good one. It's just practice!

 

But eventually TSM will age out, maybe not for years, but I want to have something to jump to that I feel equally confident with.

[NancyGormezano]

If you're on the Mac TSM2 is not there for you, that's true.

 

TSM2 still works on the current 32-bit A:M. Since you can have both 32-bit and 64 bit installed, I see that as only a minor problem for PC users.

 

My question is directed to the PC users who have it, run it and are still stopped dead in their tracks by it.

 

That normal default rig that TSM2 installs should be more than enough for to handle any animation situation that someone who is starting out with their own characters will encounter. I know it's more than enough because I've done lots of animation with it.

 

My premise is that if you have made a character, you will be animating it sooner with TSM2 (or any "installer" rig) than if you try to learn rigging on your own and make one from scratch. Most beginners who try to build a rig from scratch will build a lousy one that is lousy to animate with and then complain that the program is the problem.

 

The scripting... it was never sold as something any user would tinker with on the script level so i think it's a cool extra that it can be done AT ALL. Changing the scripts is not something a beginner needs to do at all to get TSM2 to work or to use it.

 

Face rigs. You do have to do your own solution for face rigs with TSM because face rigs aren't part of it. But I've never had trouble putting my face rigs on it. The only bone up there is the head bone, navigating that and making your face bones children of that shouldn't be a big issue.

 

Didn't know you were looking for input only from beginners, with a standard character. Or maybe just from Gerry?

 

It's not a matter of being stopped dead in one's tracks. The trouble comes for most (especially me, and I believe Holmes also expressed the same dilemma at one point) when they want to modify it. Then they have to understand the resultant overly complicated rig.

 

I don't blame A:M for this.

 

BUT I sure would like to see some real time rigging function (not via constraint, or pose, more like a "snap to") in A:M that performs similar function as the TSM script commands that allows one to create, change a bone or null to have same position (start, end points), and oriention, roll like an already existing bone, null (maybe even cps?) without having to manually do it, or input values. This could help in accurately getting control bones, nulls lined up with geometry bones, and maybe even cps.

[robcat2075]

One problem with TSM2 ( and TSM1 really) is that the documentation so briefly glossed over the the need for YOU the user to do proper CP weighting or fan boning that a lot of people missed that entirely and wondered why TSM wasn't making it look good automatically.

 

I guess the problem with TSM2 is that it is just the control rig to move the body around and not something that properly weights or fan bones an elbow or a shoulder for you.

 

We need to find a way to teach that elbow and shoulder weighting because they will need to know that no matter what they rig their character with.

[robcat2075]

Didn't know you were looking for input only from beginners, with a standard character. Or maybe just from Gerry?

 

Everyone's input is valuable so i can see the scope of the disappointment or dissatisfaction. But some of it there are ready answers for.

 

It's not a matter of being stopped dead in one's tracks. The trouble comes for most (especially me, and I believe Holmes also expressed the same dilemma at one point) when they want to modify it. Then they have to understand the resultant overly complicated rig.

 

Is it harder than building the same rig from scratch and then adding the new thing? I would think those are both hard.

 

I don't think these rigs are OVERLY complicated, I think they are what they need to be to do everything they do.

 

 

BUT I sure would like to see some real time rigging function (not via constraint, or pose, more like a "snap to") in A:M that performs similar function as the TSM script commands that allows one to create, change a bone or null to have same position (start, end points), and oriention, roll like an already existing bone, null (maybe even cps?) without having to manually do it, or input values. This could help in accurately getting control bones, nulls lined up with geometry bones, and maybe even cps.

 

You can exactly duplicate any bone in place by CTRL-dragging it in the PWS onto its parent. You will have to delete the new bone's children that also got copied if you don't want them.

[NancyGormezano]

Is it harder than building the same rig from scratch and then adding the new thing?

 

Yes it is - when you have your own rig or understand your rig then it is easy to modify. The problem with TSM is that it's hard to understand. You understand it because that's what you use, have written scripts for, have installed it, modified it many times. I found it overly complicated when I tried to analyze it. There are simpler ways to do what it does with less bones. I found it much easier to analyze and understand the 2008 rig, and the literig.

 

I don't think these rigs are OVERLY complicated, I think they are what they need to be to do everything they do.

They all do what one needs mostly. It's that some do it with WAY more bones than necessary, or have the bones organized in a way that make it hard to figure out for the user who is not intimately familiar with the how, why of the particular rig. It's a mess.

 

See comparison image. I did not include Squetch Rig - with almost 1000 bones! but also with the most functionality, but most of it never to be used, and the workings only understood by 2-3 people? And way toooo many bones show up in the timeline when animating - way too confusing for me.

 

For my comparison - Lothario (hybrid of literig, with 2008 legs) uses 197 bones and has way more functionality (body, hand rig, face rig, dynamic whiskers, dynamic tail, dynamic ears, some squetch) than the 2008 rig at 209 bones (body, hand rig only) or the TSM basic character with the most at 257 bones (body with squetch).

 

I could even trim more bones out of Lothario (ones that aren't necessary). These extraneous bones only exist because I used part of the generalized install method for Literig and liteface. The reason I would trim the number of bones down is because it simplifies animation, and sometimes these extraneous bones get keyed for no reason and complicate the workspace and confooose my widdle brain when animating.

 

BUT I sure would like to see some real time rigging function (not via constraint, or pose, more like a "snap to") in A:M that performs similar function as the TSM script commands that allows one to create, change a bone or null to have same position (start, end points), and oriention, roll like an already existing bone, null (maybe even cps?) without having to manually do it, or input values. This could help in accurately getting control bones, nulls lined up with geometry bones, and maybe even cps.

 

You can duplicate any bone in place by CTRL-dragging it in the PWS onto its parent. You will have to delete the new bone's children that also got copied if you don't want them.

 

Yes I know that - I have used that obviously when initially creating control structures. The function I am looking for is most useful when one is tweaking, modifying the geometry, moving, changing geometry bones and the control bones, relationships already exist, and you want to move the control bones to align again to the appropriate geometry bones - to realign, and get precise, rather that eyeballing or entering long strings of numbers.

Edited February 28, 2012 by NancyGormezano

[robcat2075]

BUT I sure would like to see some real time rigging function (not via constraint, or pose, more like a "snap to") in A:M that performs similar function as the TSM script commands that allows one to create, change a bone or null to have same position (start, end points), and oriention, roll like an already existing bone, null (maybe even cps?) without having to manually do it, or input values. This could help in accurately getting control bones, nulls lined up with geometry bones, and maybe even cps.

 

You can duplicate any bone in place by CTRL-dragging it in the PWS onto its parent. You will have to delete the new bone's children that also got copied if you don't want them.

 

Yes I know that...

 

Ahh... what you really meant was to create a new bone that used the start and end points of different bones? Like TSM can do.

 

 

hmmm...

[NancyGormezano]

Ahh... what you really meant was to create a new bone that used the start and end points of different bones? Like TSM can do.

 

Yes (along with orientation, roll, or aim like TSM) and as well, to repeat myself:

 

The function I am looking for is useful also when one is tweaking, modifying the geometry, moving, changing geometry bones and the control bones, relationships already exist, and you want to move the control bones to align again to the appropriate geometry bones - to realign, and get precise, rather that eyeballing or entering long strings of numbers.

 

I typically just import rigs from one character to the other, and then just start realigning the geometry bones, control bones. No Install process necessary. Just a tweaking hell.

 

Or will start with fully rigged character, modify geometry into entirely new character (so I don't have to do major weighting of cps), and then do the tweak dance. And then do the compensate dance. Would be nice to have some auto compensate (zero the offsets ?) method as well.

 

Obviously we've all been struggling along for 10-20 years without these functions.

[itsjustme]

See comparison image. I did not include Squetch Rig - with almost 1000 bones! but also with the most functionality, but most of it never to be used, and the workings only understood by 2-3 people? And way toooo many bones show up in the timeline when animating - way too confusing for me.

 

Believe it or not, there used to be more bones than that in the Squetch Rig, but I've trimmed it down considerably. Every bone has a purpose or they wouldn't have been added (there's a lot going on in there)...and the non-standard (compared to other rigs) bone parenting is what allows the rig to squetch the way it does. About the bones showing up when animating...I've tried to name non-animating bones with a "z_" at the front so that they would be at the bottom of the Timeline list...as far out of the way as possible. Could it be done better? Probably, but I think that can be said about most things.

 

As far as not many people understanding the rig, that is my fault, Nancy. I haven't gotten the tutorials done yet (I have helped anyone that asked for help or a modification for their own purposes though)...and it has been a few years. In my defense, the rig has evolved and the tutorials would have been quickly outdated. I am in the very slow process of putting the tutorials together and seeing if I can add one last function before I impose a "feature freeze"...there is a lot on my plate. I did put together a few tutorials showing how things work in the Squetch Rig, like the arm/leg rig and the installation rig (in my A:M Films tutorials...see my signature for the link), but there are quite a few more that I need to get done.

 

I think every rig designer has done their best to make a good rig, regardless of the focus. For the Squetch Rig, it has always been about squetching with a "kitchen sink" approach to capabilities...it does as much as it can in a single package. However, not everyone is going to be satisfied with what is available...that's a good reason to build another rig.

[robcat2075]

Is there a program out there that does all this better? Is there a better system out there that someone can point me to a demonstration of.

 

There are programs like Poser that make a character with a rig in it already (that's easier!) but you can't make your own character in Poser, you have to like the characters that Poser can make..

 

In Maya you manually place your bones and create your constraints one at a time like we do in A:M, and the various auto-rigging scripts and programs they have are not conceptually different or easier than what we have in things like TSM2 or the Squetch rig installer. If they were easier we'd see lots of user made characters being animated but we don't. I see a few stock characters that are widely used because most people with Maya can't rig their own characters (or maybe can't even make their own characters).

[mtpeak2]

The function I am looking for is useful also when one is tweaking, modifying the geometry, moving, changing geometry bones and the control bones, relationships already exist, and you want to move the control bones to align again to the appropriate geometry bones - to realign, and get precise, rather that eyeballing or entering long strings of numbers.

For the 2008 rig, if you readjust geometry bones, all you need to do is re-export the model from an action to realign the control rig (following the instructions for exporting as you would have from the original installation).

 

I could trim down quite a few bones in the rig, but you would lose a few feature and it wouldn't really make the installation any easier.

 

I don't know how much easier I could make installing the rig. You position half the geometry bones, use the mirrorbones plugin to mirror the left side bones and weights, then export from action to position the control rig.

[RU2D4]

I've been working through the Barry Zundel videos but keep running into a wall. I've learned a ton, but I think my time will be better spent getting back to the "Nightcallers" trailer.

 

I think I sort of understand rigging better than I did before, but in the videos he encounters occasional problems (mostly with Compensate Mode and parenting hierarchy) and it just gets my mental knickers in a twist.

 

I don't need the rigging knowledge for "Nightcallers" since I use The Setup Machine, and I am making good progress on modeling Biff, so it's back to them lovable bugheads for the nonce!

[Darkwing]

The function I am looking for is useful also when one is tweaking, modifying the geometry, moving, changing geometry bones and the control bones, relationships already exist, and you want to move the control bones to align again to the appropriate geometry bones - to realign, and get precise, rather that eyeballing or entering long strings of numbers.

For the 2008 rig, if you readjust geometry bones, all you need to do is re-export the model from an action to realign the control rig (following the instructions for exporting as you would have from the original installation).

 

I could trim down quite a few bones in the rig, but you would lose a few feature and it wouldn't really make the installation any easier.

 

I don't know how much easier I could make installing the rig. You position half the geometry bones, use the mirrorbones plugin to mirror the left side bones and weights, then export from action to position the control rig.

 

I know what I'm about to say isn't all that helpful to this discussion, but Imma throw it out there anyways, because I'm definitely the beginner that Robert refers to earlier on. First of all, about 50% at least of what you guys are talking about has gone in one ear an out the other for me basically because of what Robert (and others) have sort of said: a beginner like me has very little concept of all this stuff. Basically, this is all I know when it comes to rigging: I can add a bone. I can click the bone and click and drag around it to attach some CPs. I can turn IK on and off. I can rename the bones. Two weeks ago I learned that I can do bone weighting or whatever and it made building my face rig a lot more functional as now the CPs were weighted based on bone rotation size thingies or something and the mesh would move more smoothly than a whole conglomeration of them at once.

 

So that said, I have never built my own rig from scratch. I'm pretty sure I never will. When in bones mode, a simple right click to display all of those options that are there and I'm lost, so I do something else I know a bit about constraints from the point of view of constraining my camera to a character in the chor, or a prop or something, but that's it. I can't do stuff with knees always pointing a direction, or feet staying on the ground while you move the hips and all that jazz.

 

So if I have a point, it would be a testimonial more than anything. I've only ever used the 2008 rig. I use it every time. It's easy to use, the instructions are clear as day and I only ever have to do half of it! It even has cool slidey things for the fingers so that they can bend! As a beginner, the 2008 rig does everything and more than I could ever hope to achieve by building a rig. Of course the day's probably gonna come where I'll want to rig something other than a human. If that day comes, pretty sure you're all gonna know...

 

I understand this discussion is really geared more for those intermediate users who are above my level of ability and want to modify pre-existing rigs and to do so, want/need a better understanding of what everything does and why, but like I said, this isn't probably helpful to the discussion at all but at least it's 100% from the POV of a beginner when it comes to rigging.

[NancyGormezano]

Mark, David - my preferences for what a rig does, and how to install it are just that...my preferences. In no way was I critiquing what either of you did. You both did extraordinarily terrific jobs, without which I would not have been able to steal your ideas, as well as from Holmes and countless others. You two in particular have both been incredibly supportive to anyone who has tried to install either of your rigs.

 

I prefer to understand what I'm doing, and to pick & choose the functionality. I prefer to be able to understand it enough so that I can fix the problems myself, rather than needing to run to someone else. People come and go. Rigs, software gets abandonned.

 

I loved animating with Squetch ver1 on TWO - it did everything I needed. But there was no way I could tweak it for anything other than adding a branch of bones for something independent of basic rig. There was no way I could have dragged a completed rig to a new model, and moved bones around to fit some new geometry. It seemed to have less extraneous bones showing up in timeline than vers 2 (SO). The extraneous bones that show up are usually related to when switching modes of how something operates. Even tho it's necessary for the rig to work, it does make animating less fluid and spontaneous when one has to look at a long list and think.... It interrupts the flow. It's a preference of mine to have minimal bones with channels showing up in the timeline.

 

If any rig did everything I ever needed and it never had any hitches in installation, and it had a minimal amount of controls that show up in the timeline for animating, and ... then I must have died and gone to heaven. Obviously that's almost impossible...I don't care how perfect either of you are. Even my rig has stooopid bones that show up.

 

I was only commenting on how easy or hard it was to analyze a rig, from a user point of view, as well as animate with it. The less bones the better. Sometimes less is more. Especially for new user.

 

User interface is all important. And the most difficult to develop. I have no experience with rigging in any other program, and I highly doubt that it would be easier to modify and understand rigging in those programs as well.

 

It is imperative for anyone who is not intimately familiar with the development process for any rig developed by someone else, to get a top down "management summary view" first , then they can progress to understanding the details. This is not easy to do in A:M with any rig that one hasn't developed, or is a beginner. It is especially difficult to analyze any rig in A:M when one has to scroll thru endless bones opening and closing branches, with bones that have incredibly long names, that appear too similar to other bone names. Or have funny conventions (like TSM). It is especially difficult to keep in ones brain all the properties, relationships, etc that influence the bones when one doesn't understand the methodology or constraint concepts used in a particular rig. They're all different. It's like reading code without any comments, documentation, and without being able to easily see it in recognizable chunks.

 

The 2001 rig was the easiest to understand -as presented in the TAOAM. No installation rig, obviously lacking in some functionality, easiest of all rigs to modify, tailor, tweak, and retweak.

 

I was able to analyze the 2008 rig and break it down into it's component parts, and save those out as separate models (eg legs, arms, spine along with their constraints), that I can import easily into any model. Same with literig. I found that I had to actually print out the complete bone hierarchy and it made it easier for me to see how things were structured in order to break it apart with respective pose/relationships, to make it more of a modular installation choice.

 

The 2008 instructions were extremely CLEAR, accurate, well done. All 75 pages. Yikes. To a user, seeing 75 pages (again I had to print it out first), that list of instructions is daunting and gives the new user the willies. Obviously one just plows thru, and it turns out to be no big deal.

 

For the 2008 rig, if you readjust geometry bones, all you need to do is re-export the model from an action to realign the control rig (following the instructions for exporting as you would have from the original installation).

 

I have found problems with exporting from an action after texturing and adding, grooming hair. It didn't always work. Decals get messed, hair looses grooming. I shy away from that now with tweaking already textured, hairified models.

 

I have also found working with mirroring bones, weights sometimes wonky. My models aren't on purpose always symetrical. At some point it becomes easier for me to do by hand than to rely on "black box"

[robcat2075]

I have found problems with exporting from an action after texturing and adding, grooming hair. It didn't always work. Decals get messed, hair looses grooming. I shy away from that now with tweaking already textured, hairified models.

 

If you could make a sample case out of that to study and post it... It would be useful to see.

[Darkwing]

I've had the same thing. Hair is very finicky to maintain groomed. In most cases it seems like it will only stay groomed in the original model groomed in. I'm pretty sure this works for importing groomed models into another model as well.

 

EDIT: And of course the time I go and do it to show that it happens and it works fine...

[mtpeak2]

I have found problems with exporting from an action after texturing and adding, grooming hair. It didn't always work. Decals get messed, hair looses grooming. I shy away from that now with tweaking already textured, hairified models.

I haven't had any issues with this in v16b. v16a had an issue with exporting materials (hair was one of the materials) which I reported and was fixed in v16b.

[Darkwing]

Oh that could be it! I only just recently upgraded from 16a to 16b and come to think of it, it happened while I using 16a

[robcat2075]

One thing I learned about TSM2 by looking at the scripts...It's not one big thing, it's many small things that are mostly independent of each other and feasible to understand. The fingers and the arms are separate things and both are fairly simple in themselves. Likewise with any of the other components. The IK Spine is the most elaborate part.

[HomeSlice]

..... My major complaint of all generalized INSTALL methods (lite rig, 2008, Squetch all are guilty, some more so than others). Too many bones in there just to make installatiion process work. Too complicated to tweak once rig installed.

I'm pretty sure everyone who has ever created an INSTALL rig has bumped up against this. All the rigs share a similar basic philosophy; the user adds the geometry bones and the installation process adds the control bones/constraints/relationships etc. The problem I ran into with the LiteRig is that I had to add a bunch of extra bones just so the install process could automatically place the control bones in the correct positions and add the constraints/relationships without user intervention. I believe anything which adds an "automation" layer to a manual process will complicate the resulting product significantly. I have the same issues as Nancy regarding analysis, customization and modification of INSTALL rigs.

 

I then tried to create a series of drag-drop modules for different types of bone structures, but that didn't pan out either. Grasshopper in SO was rigged with this approach and even though it was possible to animate him reasonably well, the rig definitely had issues.

 

That's why I pretty much build every rig from scratch now.

 

 

We need to find a way to teach that elbow and shoulder weighting because they will need to know that no matter what they rig their character with.

It is difficult to break the process of weighting a complicated area down into a methodical set of steps. When I was planning out my Face Weighting tutorial ( http://www.hash.com/forums/index.php?showtopic=34473 ) I re-rigged the same face mesh 6 or 7 times while keeping notes in order to try to figure out a no-brainer set of steps, but in the end I didn't quite achieve my goal. The user still has to understand a little about the process and philosophy of weighting in order to apply the concepts covered in my tutorial to his/her own models.

 

What might work better is a simple tutorial with a 100x100 grid. Place a bone at the center CP on the grid and assign the central CP to the bone. Then guide the user through the process of making all the other CPs in the grid move in a fluid way in response to the rotation of the bone. That pretty much gives the user all he/she needs to know.

 

 

----------------

If we had a javascript or python interface to the sdk (and decent documentation on how to use it) we *might* be able to create scripts to install animation rigs without adding that extra layer of complexity to the resulting rig, but I'm not going to spend the next five years learning C++ just to create an install script when I can rig my own character in a couple of afternoons. Actually, I would LOVE a way to interface with the sdk using AutoHotkey. AutoHotkey is by far the easiest way to create simple Windows desktop applications. AutoHotkey_L will also interface with Windows COM objects. But there does not seem to be a viable way to run complex AutoHotkey scripts on a Mac. The version for Mac (IronAHK, seems to have stopped development).

 

However, having said that, TSM has a scripting environment and that rig still seems pretty complicated when I try to modify it to suit my own wants/needs....

---------------------

 

In the end, it seems the best solution is a series of tuts to teach the user how to build his/her own rigs.

1) basic humanoid rig (with IK/FK arms and legs)

2) hand gizmo

3) foot gizmo

4) face rig

5) squetch components

6) quadruped rig

7) insect rig

[itsjustme]

The problem I ran into with the LiteRig is that I had to add a bunch of extra bones just so the install process could automatically place the control bones in the correct positions and add the constraints/relationships without user intervention. I believe anything which adds an "automation" layer to a manual process will complicate the resulting product significantly. I have the same issues as Nancy regarding analysis, customization and modification of INSTALL rigs.

 

If the installation bones have "INSTALL" in their name, you can run the InstallRig plugin to have them cleaned out.

 

This does remind me to mention why some of the "extra" bones are in the Squetch Rig. There are bones used purely for storing measurements/numbers for calculation in Expressions (see A Scalene Arm Rig as an example), nulls added as controls (for the FACE controls, on-character switches, etc), bones added so that the user doesn't have to manually reset the compensates (there used to be around 80 compensates that needed to be reset during installation), a version of control bones for every setting (a set for FK, a set for IK and any other setting...everything has multiple settings), discrete control and geometry bones (a basic rig would generally have one set for both in a lot of places) and (of course) bones that facilitate squetching (which a basic rig would exclude).

 

On the subject of automated CP Weighting...any automated method would require a very simple rig in order to be accurate. I don't think the Squetch Rig could take advantage of an automated weighting method since there are multiple instances of co-located bones. The best way to "ball park" the CP Weighting with the Squetch Rig (although it would work with other rigs as well) would probably be by using the Transfer AW plugin (using an already weighted character to transfer to the new character). I'm going to release a generic character (in addition to Squetchy Sam) that might be helpful for this purpose.

[Darkwing]

Y'know this is all very interesting. It's very neat to see how essentially the "same" task has so many different ways of being accomplished. It's really quite a diverse topic, rigging that is, isn't it?

[robcat2075]

There's a reason that big studios have to actually advertise for a "TD" , a person who rigs, and not for "animators".

 

People who "get" rigging are much rarer than animators.