Jump to content
Hash, Inc. Forums

Squetch Rig Setup Machine


Dhar

Recommended Posts

Now that the new & improved squetch rig is nearing completion, I've been thinking (more like day dreaming really) of ways to make installing the rig as easy as possible, especially when it coms to CP assignments and weighting.

 

My thought is to have some sort of standard when modeling the character. For example; to model a simple cartoon character like Sam, there need to be parameters of splineage that will facilitate the installation of the squetch rig. A forearm (for example) would have no more than 6 circular splines that correspond with the squetchrig's forearm bones, then, with a single Action (click & drop), the forearm CPs (and the rest of the model) get assigned and weighted in one go.

 

Is such an idea even possible?

 

 

::::: you may say I'm a dreamer - but I'm not the only one:::::

Link to comment
Share on other sites

you may say I'm a dreamer - but I'm not the only one:::::

 

 

OHNO! You split-up the Beatles!!!

 

SOUNDS feasible... why not just provide a basic model (Thom?) with all the appropriate numbers of CP's and just let the end-user move them about to make the character of their dreams....

Link to comment
Share on other sites

I did some experiments that might do the job using the "Transfer_AW" plugin...if you do a search, you should find it in the forums here. The plugin is on Steffen Gross' site, here.

 

Hope that helps, Dhar.

 

 

 

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

EDIT

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

 

Here is the post where I give the results of my experiments with "Transfer_AW".

Link to comment
Share on other sites

hmmm, for me weighting cps is the fun-part of rigging, you weight and test and weight and test.... and the caracter moves smoother and smoother... can´t really explain why, but i love that... others spend an evening with sudoku, i open up a bottle of red wine ad weight cps.. ;)

 

 

but i asked myself lately if it would be possible (i´m not a programmer at all, so i´m just guessing) to write a plug-in for resetting the compensates... that´s the pain-in-the-ass-incredibly-boring-part for me....

Link to comment
Share on other sites

SOUNDS feasible... why not just provide a basic model (Thom?) with all the appropriate numbers of CP's and just let the end-user move them about to make the character of their dreams....

 

Hmm, that sounds similiar to poly modeling, but its one option.

Link to comment
Share on other sites

I did some experiments that might do the job using the "Transfer_AW" plugin...if you do a search, you should find it in the forums here. The plugin is on Steffen Gross' site, here.

 

Hope that helps, Dhar.

 

 

 

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

EDIT

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

 

Here is the post where I give the results of my experiments with "Transfer_AW".

 

Yeah, it does help, however, I was hoping for a more complete install rather than "close" and then having to tweak the tight areas.

 

What I was aiming for is some standard method of modeling that would facilitate the squetchrig in such a way that no further CP weighting would be necessary. Refer to my pic, you'll see how some splines on my model correspond with Sam's (red line). But, because I was following Barry Zubdell's method of modeling, many splines do not correspond anymore, which forced me to individually (painstakingly) assign and weight the CPs to achieve smooth transitions.

 

My thinking is to somehow find a standard number of splines (or way of modeling) that will give the best performance. You told me when I was modeling Gumbo that no less than 8 splines should exist for the eyelids to perform satisfactorily. Now if there was a way that an Auto_assign would exactly assign and weight those 8 splines, that would eliminate the need of tweaking.

 

You see, now that I have some experience with the Squetchrig I am mor inclined to re-model Gumbo in a fashion that will easily accept the Squetchrig without any after-tweaking.

 

Is that possible?

post-5367-1206314365_thumb.jpg

Link to comment
Share on other sites

hmmm, for me weighting cps is the fun-part of rigging, you weight and test and weight and test.... and the caracter moves smoother and smoother... can´t really explain why, but i love that... others spend an evening with sudoku, i open up a bottle of red wine ad weight cps.. ;)

 

 

but i asked myself lately if it would be possible (i´m not a programmer at all, so i´m just guessing) to write a plug-in for resetting the compensates... that´s the pain-in-the-ass-incredibly-boring-part for me....

 

OK, the next time I model a character I'll give it to you CP weight it ;)

 

For me, I want to animate, NOW. Which is why I have little patience for modeling and rigging.

Link to comment
Share on other sites

So, you're saying you want to use Sam as the base mesh? But even doing that, and manipulating especially his face into a new character.......still, he'd need different cp weighting. Why not try using the bone falloff property to visually get the weights. It mightn't be perfect first time, but at least, it will give you the right bones in the weighting list.

Link to comment
Share on other sites

Come on people. You're willing to learning everything else about AM, but not rigging. Give me a break. Rigging is no different than learning anything else in AM. You may be able to model a character, but if you can't rig it, it isn't worth much.

Link to comment
Share on other sites

I don't think it's so much the learning as just the sheer size of the job. Since I've been doing a lot of it lately, I've started to find some shortcuts, but it's still a pretty big task in front of you when you start going at it.

 

It is a pretty satisfying feeling, though, when you get it done.

Link to comment
Share on other sites

Size of the job? That's nothing compared to building models, texturing, sets, lighting, animating and so on. Rigging is only a small piece of the puzzle.

 

I'm not much of a modeler and I don't claim to be. The reason for this is because I never sat down and was determined to do so. But, I'm not going to sit there and complain that modeling is to hard and I need an auto modeler because I don't have the experience.

Link to comment
Share on other sites

  • Hash Fellow

I can imagine a program that would look at specially named groups you had created and look at the location of the CPs in them and then weight them semi-correctly, but it would presume a substantial similarity in function and appearance to an "ideal". And the specially named groups would be so elaborate that you might as well just weight the CPs instead.

 

I think well done bone fall-offs might be the best quick and dirty CP weighting compromise.

 

Come on people. You're willing to learning everything else about AM, but not rigging. Give me a break. Rigging is no different than learning anything else in AM. You may be able to model a character, but if you can't rig it, it isn't worth much.

 

I agree. If you took the time to put all the CPs in the right place to make your character you can take the time to weight them.

Link to comment
Share on other sites

I agree with Robert all the way, in that you may as well weight the model yourself. If you can create a group for a bone to be assigned to and weight, you may as well just do it.

 

Bone fall-off will only get you to a certain point. All models are different and setting a default for fall-off will, for the most part, be a waste of time. Fat arms, skinny legs vs. skinny arms and fat legs, it just doesn't work. Half the bones in any rig are too close for fall-off to work correctly anyway. So, if you have to adjust the fall-off for each model, you may as well just manually set the weighting, you're gonna have to fine tune the weighting anyway.

Link to comment
Share on other sites

I think Robert expressed it better what I was talking about. A set of parameters of specifically named group of CPs or splines that the Squetchrig would recognize and assign the CPs to the corresponding bones AFTER the rig has been sized to your model. So, it doesn't have to be Sam or any specific model. What I'm after, I guess, is some form of standardization of modeling - CPs, groups, splines- that will aid in the ease of the rig installation.

 

You see, what we have in the Squetch rig is a standard of bones and actions that are designed to give top notch animation. But if the model (such as Gumbo) was not modeled with the Squetchrig in mind then I have either limited some of the features of the rig, and/or overwhelmed myself with, as lergento mentioned, a sizable project. The model and the rig need to meet somewhere in the middle. Now that we have the rig, I feel there need to be specific way of modeling that will utilize the most out of the rig.

 

Did I confuse the issue further??? :huh:

Link to comment
Share on other sites

What about it? Are you still creating a group for that? Or do you want this done magicly?

 

When you animate, do you just key pose? Do you want a plug-in to set the inbetweens? What about cleanup on your animation, do you want that done magicly too? What would be the point of it all then, if you have all these things done for you.

 

If you don't want to learn how to install a rig, that's your choice. But, if you put the amount of effort into rigging that you do to learn how to animate, you would be able to install the squetch rig by now.

Link to comment
Share on other sites

  • Hash Fellow

I think if bone falloffs had an onscreen adjustment GUI (instead of typing in numbers into a box) they'd be more accessible. but even some defaults would at least get your character to where he can walk around and turn his head and scratch his butt. It would be quick gratification.

 

"Compute All CP weights " seems to recognize which bones and CPS are visible so I can sort of imagine creating a few mega groups that isolate CPS by their function. Say "regular bones" and "fan bones" and do the COmpute in a separate pass for each and tha t would eliminate a lot of slop that a one pass compute would have.

 

 

But still, not a finely done job.

 

And Faces are still a complex mess.

Link to comment
Share on other sites

What about it? Are you still creating a group for that? Or do you want this done magicly?

 

When you animate, do you just key pose? Do you want a plug-in to set the inbetweens? What about cleanup on your animation, do you want that done magicly too? What would be the point of it all then, if you have all these things done for you.

 

If you don't want to learn how to install a rig, that's your choice. But, if you put the amount of effort into rigging that you do to learn how to animate, you would be able to install the squetch rig by now.

 

Your points are valid. But weighting each CP and then changing the weight in trial & error just seems like a lot of work. Inbetweening was a lot of work in the old days of cell animation, which is something the computer can do now. I'm asking if there is a way where the computer can do the weighting than having to do it manually. There obviously is, but to minimize the cleanup, I think a model with splines stitched a certain way can make the job easier. That's what I'm looking for.

Link to comment
Share on other sites

  • Hash Fellow
Inbetweening was a lot of work in the old days of cell animation, which is something the computer can do now. I'm asking if there is a way where the computer can do the weighting than having to do it manually.

 

It's called "Compute all CP weights". It aint' great but neither is computer inbetweening. If you want great animation, you have to adjust a whole lot of keyframes yourself. If you want great CPweighting, you have to adjust a lot of weights yourself.

 

 

In the polygon world they have "Weight maps", but those don't paint themselves either.

Link to comment
Share on other sites

this could only work if you have very similar characters with almost the same proportions. imagine you have to assign all your cps to groups, and then the "automatic" weighting will be done, if you have a character with a slightly different geometry you will have to correct the weights in a bunch of places.

i think the amount of time you would spend on that would be almost the same as assigning and weighting every cp by hand...

Link to comment
Share on other sites

"Technology doesn't make the motion picture, people do. You're not an animator just because you can move an object from point A to point B. (You're) someone who breathes life into a character, which is something the software and technology can't give you."[/quote]

 

This also applies to the rigging that takes place before trying to animate.

Link to comment
Share on other sites

True, but something like this Pinocchio application would be...

 

http://www.mit.edu/~ibaran/autorig/

 

It's designed for children and amateur animators and I could see the benefits and limitations of such a system. Granted, a setup as complex as the squetch rig would probably be impossible to adapt this to, but for folks who just want to make their own characters and move them around, it would be a pretty cool thing. Especially since it does the rigging and weighting in under a minute!

 

The video is interesting.

Link to comment
Share on other sites

There are ways of doing it. It only takes someone with the vision, the programming ability and the drive to make it happen.

 

I think that we will see more and more automation coming into the rigging process in 3D applications over the next few years but learning to build your own rigs can be very engrossing. It is a challenge and it is very satisfying when it all comes together but for the average person who got into this in order to animate it can be an unwanted and frustrating distraction.

 

The artist is the customer and the customer is always right. ;)

Link to comment
Share on other sites

This is interesting Mark, but this is just a geometry skeleton with motion capture applied to it. At least that's the impression I get from it. I don't think you can manually animate it.

 

A nice feature idea would be an auto fall-off. The fall-off would detect the surrounding geometry and set itself accordingly. Then you wouldn't have to adjust the parameters of the fall-off for each character you install a rig to. This still wouldn't help in certain situations, like the face, but it would work alot better than it does now.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...