CA Tools We Want


#1

Ngrava said in another thread…“Character Animation Tools: What tools?! I’ve seen a few suggestions here and there but really guys… Be more specific then “Better” and “More”. I like the idea of preset joint types. How about dynamic joint deformers? Define an area (cluster) around a joint, attach a dynamic deformer to it, and specify some simple parameters for bulging, skin thickness, and muscle mass and so forth, then just start animating and it figures out how the joint should look. This is kind of like a muscle system but actually quite simpler. No need for joint driven morphs, this is real smart skin! How about collision detection added to softIK? How about a whole, real-time dynamics system that allows you to push and pull things around like in the real world while you’re animating? How about spline based IK chains for noodle arms on cartoon characters? And, cartoon bones that stretch and compress automatically? I’d also love to see a pin based IK system like Motion Builder’s in Cinema. Let’s get more creative!”

I agree…I see many requests for better and more but not alot of detailed requests… So here is a list of some of the tools that could be implemented …

  1. Preset joint types -single rotation plane hinge joint and dual rotation plane hinge joint.

  2. Dynamic joint deformation for smart muscle and skin simulation

  3. Collision detection based on deformed mesh surface normals, where the mesh clamps the bone if it detects a collision and stops penetrating objects automatically.

  4. Real time dynamics. Tell the rig how many units it weighs and it assesses the bone and mesh connections for volumes of mass and rigidity to determine the weight distribution over the character, with big tyrannosaurus thigh and belly showing the weight when it moves because the volume around the bone is much larger than the arms. Smart skin handles the deformation.

  5. Cartoon bones. These would scale down the x and y axis as the z axis is scaled up. Pick your scaling parameters and then rig and rock.

  6. Facial rigging… There are a few things every face has…a mouth an eyes. The eyes blink and have a few positions for opening and closing. The mouth opens and closes and the chances are that it will use phonemes. These can be parametric to set up with selection sets and dropping the selection tag into the controller interface. Type in “M” and the lips purse. Type in “O” and the lips make an O" shape. Use voice recognition to lip synch your mouth automatically.

  7. Learned or procedural behaviors. An animation routine should be able to be packaged and dragged and dropped onto any other rig you build.

  8. AI. OK… I am psyched about this Endorphin technology. The exponential benefits of being able to share evolved taught behaviour to a rig amongst in house animators or the community at large is boggling. Some rigs will become the Jackie Chan or Jhett Li of fighting martial arts style. Another rig may be the best dancer ever…just put some music on the timeline and the rig performs… tell it the hip grind sux and it is no longer part of the routine. You could have dejected rigs, confident rigs, drunken rigs, expressive rigs. You could have a “suitcase” full of rigs and hire yourself out as the guy who can best make a CG character do stunt fighting, while continually training your rig in new Kung Fu and Aikido moves for example… When a punch is thrown at it it knows ten ways to deal with it and choses the best, most effective and simplest method to block and counter the punch. FWICR Endorphin AI technologies were used in LOTR for the battle scenes… It is here and it needs to be implemented.

If we want character animation tools I think we need to be specific so the great programmers at Maxon can get a grip on what we need. We do not need to catch up… We need to revolutionize CA setup and acting. Feel free to add to this list or add suggestions as to implementation of any of the tools mentioned already. Adam said that we talk and talk about animation but rarely see it…or something to that effect. I would think that it s because we are close to being perfectionists and we want to put outstanding work out there. So far most of my CA time has been put into experimenting with rigs that won’t break and whose feet clamp to the floor without unwanted deformation or null target slippage during simple foot rolls. It would be nice to be free of the technical setup aspect and tagging or applying Xpresso to every bone to just get it to act basically without breakage and concentrate on making the rig act and move dynamically.

best
BT


#2

Plus:

http://www.cgtalk.com/showthread.php?t=182217


#3

I think I would try to keep this all in the thread that michaeli pointed out. Especially since it was started by Kai.


#4

Why? There are a zillion threads on GI and rendering and neverending Poser dialogues… What can it hurt to have further discussions on a new thread about CA in C4D? There are always new readers coming on and the inclination is to not dig back but read the front page… Not everybody reads this forum daily. And i want to discuss this further… This is what this forum is for… Correct me if i am wrong.

The other thread had some very excellent ideas in it as well…but it was getting cumbersome as evidenced by the folks who chimed in saying they had not had the chance to go through the whole thread… I am sure Kai is a nice guy and i know he is knowledgable but I don’t play follow the leader. I am more interested that Maxon comes through on the CA front and to keep posts on CA at the front of this most frequented forum will let them know it is an important issue. As someone said… 3D CA is the Holy Grail. I agree.

best
BT


#5

I guess it’s not a question of following the leader down well-throdden corridors and into the valley of steel, but of keeping good ideas together.


#6

Alrighty then…didn’t mean to get you all riled up. I’m right there with you with CA being the Holy Grail, just figured if everything was in one place it would be easier to compile a complete list. When this list gets long and cumbersome, are you going to start another one?

Maybe is should be made a sticky or just have someone compile one list at some point.

Again…sorry… :shrug:


#7

That thread was compiled into a list and submitted, feel free to get this one going. I dont’ want to see this come up every week like rendering ones do, but it the discussion comes up once in a while its good to remind maxon that there are people who feel other things besides rendering are important to them.

Myself, I’ll chime in and say before anymore CA stuff I’d like to see maxon focus on the A itself. no matter how sweet of a rig you build (and honestly with mocca II and the CDIK and other plugins available you can build some very impressive rigs now) Animating can still be a really problem in cinema. our timeslider is pathetic compared to max, maya, LW, and XSI. They all for the most part have a comparable list of features int heir timesliders (which is what they call their timeline) that our timeslider isn’t capable of.

Our dopesheet is seriously outdated without any major changes since version 5 at least.

and our Fcurves, currently the best of our three editors, is missing one core part of its functionality, applying functions to the curves, we currently can only modify the actual keys on the fcurve, while apps like maya have breakdown keys that you can edit in fcurve and change automatically when the true keys are adjusted, as well as the ability to apply deformers or equations to change the actual curve, without destroying the original keys.


#8

Great points Kai! Seeing an overhaul on that level is really the start

ed


#9

…what dopesheet???
:wink:


#10

well its a matter of terminology really, our timeline manager is essentially a dope sheet, and our timeslider is essentially a timeline if you use other applicatiosn terminology.


#11

Hi, Kai, have you tried CAT (http://www.charanitech.com), I found it incredibly easy to use and powerful, should C4D’s CA part follow that way?


#12

Hi guys… Thanks for carrying it on… I agree with the timeline issue. I haven’t had alot of experience with Maya and XSI, but I do know that when i start to animate a segment that I get overwhelmed clicking through the timeline and scrolling down it.So I go and rig another character or model a further set piece. Apparently there are better solutions to this and I am all for anything that helps translate my imagination to CA acting.

On other integral tools I note that i have to pot a shoulder bone and bicep roll bone on the upper arm to stop shoulder deformation on upper arm roll. The same occurs on the forearm where three bones take a partial rotation to the wrist bone so the forearm deforms properly. A Twister Bone would suffice as one bone where two or three are needed. The bone should have falloff from the root so twist can occur from tip back or root forwards. Then I can just weight this bone and I have proper twist on that limb section. This would also work on tentacles and similar appendages.

maybe a sticky should be made but I think we should just bloat this one out and next week after this leaves the front page, replace it with further ideas, gripes, suggestions and keep it up till we have what we need to get the job done with C4D… I think you will see alot of independent one man studio shorts that will light some socks on fire.

Best
BT


#13

The Char tools I want can be summed up in a word: Bonderland.

I wish someone would pick up the ball and finish this or Maxon include tools like these in the next version.

A lot of Bonderland functionality is provided by Cactus Dans IK tools. But Bonderland has a lot of other handy tools that should really be in Mocca. Maxon did include the late Klaus’s Tag Manager in R9. So I don’t see why they can’t work something out with Bonderland.
:shrug:


#14

regarding bonderland thats up to ideagraphics.

thing I like about bonderland and CDK is tha they are about the core tools and not the auto rig.

I know a lot of people want autorigging, and most of you know I don’t. Thing is its not that i don’t want a system to rig my character more automatically, its that autorigs work the way the designer made them and not how I or you should necessarily want to work. CAT is a prime example. Nice system, but where there controls are located and such is quite lame and typical of max fashion. It is an impressive system, and setting up skeletons is quite nice in CAT, I’d take it any day over motionbuilder, but I’d rather see maxon, make the core tools properly, and then if people want autorigs that bad, leave it to the rigging guys to make them. There are so many auto rig systems for maya, because making a rig into an autorig isn’t too difficult of a process. This is where I’d like to see maxon go, make the core tools right, make it easier for me to make an autorig (right now with some expresso i can make it easy to drop in secitons of a rig connect them and then fix it, but still not exactly automatic is it?)


#15

It is getting easier to rig with experience. I get to know each systems limits and strength. I do not want autorigs ether…but a few autotwist/forearm or shoulder/bicep bones and autojointed knee/finger type bones would be nice time saving tools. But I find that if you make a rig that walks well with IK feet you cannot knock him about or have him do gymnastics very easily without keyframing the bejeezus out of the feet controllers… It is little things like this that baffle me and make me smoke too many cigarettes staring at the rig and hierarchy and tags and Xpresso nodes by the dozens…And forget about working foot roll…I am trying to get five clawed tiger paw and eagle talon claw roll. aaaaarrrrgggghhhhh!

Best
BT


#16

About CAT. There’s some really stupid limitations that it has. If you set it up so that you pose your legs using IK, you can’t control your knee position. Meaning your knee will ALWAYS point forward, regardless of how your feet are positioned. If you use FK however, this isn’t a problem. Maybe I just missed something, but that really started bugging me when I was using the trial.

I’m still looking forward to seeing what CAT 2 is all about, not to mention Cat Muscle.

Have any of you guys tried Messiah? Looks pretty cool to me. Maybe Maxon could steal some features from it.


#17

The knee controller/IKRP solver should go in the foot so the rotation plane of the knee follows the pointing of the foot. Does CAT not allow something this elementary? I mean placing your constraints and goals inside the hierarchy so they are portable and the rig reacts automatically to other parts of the rigs motion.

Best
BT


#18

CAT doesn’t allow any kind of rig controls. You have to select the bones directly. As far as the knee goes, you have to select the upper leg bone (what’s that called again?) and adjust the swivel angle. For some reason I could never animate the swivel angle, but it was surely a bug or something. I’ve seen a lot of people on the CAT forums asking for different types of controls, so hopefully CAT 2 should have some kind of controls. From what I’ve read, CAT 2 should be out in the next month or two, as well as CAT muscle, a muscle simulation app.


#19

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.