Smart bones change everything! Some examples

[funksmaname]

you're right... right now it doesn't do any of this - they only work on rotation and can only control point motion, curvature and line width - but support for more funcationlity will be built in in future I expect.

however, MorphDials are built on single frame actions, so there is no inbetweens available as such. With smart bones, you can have as many in betweens as you want on a single bone, so essentially you can put every phenome on a single bone to mix them in rotation... I've not too heavily experimented with this yet, but you could probably use this in combination with step or new 'intervals' to make pretty good lipsync (I will experiment).

Perhaps proximity control could be added later or through scripting - I'm not a scripter, or know anything about whats available or not, but perhaps the future of MorphDials is to build on smart bones through scripting to add new ways to use them? It would probably result in better performance and less workload for Rudiger I would imagine/hope.

[eago]

This bones look really cool
Was wondering how complex is to get one of those rigs done? How long does it take you the get one of those characters rigged? hours? days? weeks?

In all videos I've seen the timeline didn't seem to change much? still only one layer at a time?

[heyvern]

The time to do these types of rigs is only determined by how long it takes you to... move the points around.

Basically these are simply actions. You just rotate your bone and move the points on the layers effected by it.

I was AMAZED at how FAST Selgin got the hang of this new feature. If I didn't know any better I would have thought he had some kind of time machine and had been using it for weeks instead of days.

-vern

[funksmaname]

yeah, although the zombie character may well have already been drawn (but maybe not) Selgin was already doing the foot turning thing on like day 2 of smart bones being included... it really dropped my jaw when I first saw that! it not only takes less time to rig than before, it would take far less time to animate as you're not constantly wrestling with bones or points that serve no purpose but to maintain the characters shape...

But what Vern says is right. If it takes you weeks to do what Selgin can do in a few hours, then that's how long it'll take

[Dodgy]

you're right... right now it doesn't do any of this - they only work on rotation and can only control point motion, curvature and line width - but support for more funcationlity will be built in in future I expect.

It would be nice, but who knows when that'll happen

however, MorphDials are built on single frame actions, so there is no inbetweens available as such. With smart bones, you can have as many in betweens as you want on a single bone, so essentially you can put every phenome on a single bone to mix them in rotation... I've not too heavily experimented with this yet, but you could probably use this in combination with step or new 'intervals' to make pretty good lipsync (I will experiment).

That wouldn't work. As the bone turns, it'll morph through every phoneme to get to the one you want. You need that circle relationship so you can go from one to the other without going through the others.

Perhaps proximity control could be added later or through scripting - I'm not a scripter, or know anything about whats available or not, but perhaps the future of MorphDials is to build on smart bones through scripting to add new ways to use them? It would probably result in better performance and less workload for Rudiger I would imagine/hope.

Its limitations certainly won't stop my buying 9, my complaints about 8 were there weren't enough high end features for pro users. This function is worth the upgrade alone. Hopefully it'll speed up my using morphdials as I won't have to use them for joints and turning like I did before, and can just use them for phoneme targets if not using smart bones as a lever for each individual phoneme and basically replacing the Blend panel.

[funksmaname]

Absolutely. Morph dials definitely act more reliably when mixing poses on similar points. As mike has pointed out smart bones get a little dumb when more than one acts on the same points and to avoid complications they should only act on unique points as they are additive. Having morph targets, and setting up phenome rings with morph dials will let MDs focus on what their best at (facial rigging) while leaving the other stuff to smart bones.

[Rudiger]

Absolutely. Morph dials definitely act more reliably when mixing poses on similar points. As mike has pointed out smart bones get a little dumb when more than one acts on the same points and to avoid complications they should only act on unique points as they are additive. Having morph targets, and setting up phenome rings with morph dials will let MDs focus on what their best at (facial rigging) while leaving the other stuff to smart bones.

Every script I've ever written for AS was done with the hope that it would some day be incorporated into the main program, so I couldn't be happier with the fact that MorphDials is becoming more and more redundant!

Also, from looking at Selgin's demos, I think SmartBones handles blending of multiple smart bones controlling the same points absolutely brilliantly (eg changing expressions with 1 bone, then changing rotation with another). Surely it's using the same relative morphing method that is available in the Blend Morphs window. That makes it pretty equivalent to MorphDials, imo. All it needs is to affect more channels and be controlled by bone position and scale instead of just rotation. If AS9 turns out to be really popular, I can see these features being added for version 9.1!

[funksmaname]

Hey dude,
I'm glad you're not frustrated by the prospect! I just tried using multiple bones for facial rigging and the result was pretty good (see 'kurmi' in share your work) and did a test with morph dials on the extreme skirt situation and the results were the same as smart bones and blend morphs, so you're right (of course).

In the situation of two legs up making a skirt fly away, it probably makes sense to just have a single smart bone for the skirt alone so you can bend it as necessary when both legs are forward, or as previously suggested, have a rarely used 'fix' bone to counter act the odd additive behaviour.

It seems I overstated the limitations of smart bones compared to morph dials earlier, they do seem to act in the same way, just less channels and comptroller methods are currently available to smart bones, which will hopefully be rectified in updates

[Yosemite Sam]

Selgin,

You are truly inspirational to me. Every time I see your work I remember why I fell in love with animation. Thanks!

[chucky]

Link Smart Bone Chain better quality

You can't create a smart action with more than one bone being 'smart' at a time , but you can do each action with the next bone's extreme positions in place. This will help the overall design of actions when used in conjunction with other actions in the mainline.
I call it a 'smart bone chain' , or if you prefer a more accurate terminology try ' compound boning' hey call it what you like but it can help to use this technique, be sure though it's about the way you position the points in this case not the program doing everything for you.

You might notice there aren't any angle constraint on the early version, that does help too but is not the sole reason that the IK works better.

In the future maybe smart bones may evolve to allow an even more sophisticated calculation which will include more than one bone, layer motion and child control.
Even without this there are many great possibilities as you can still use the old action techniques for layers within groups.

[madrobot]

very nice mate. i will be in touch re: using all your ideas for free.

[Distinct Sun]

Minds will be ( or just have been) blown! Great stuff! thanks for sharing guys!

[RichardU]

I woke up thinking about how Smart Bones changes layer strategies. I am already accustomed to point binding, and rarely use layer binding. With Smart Bones, I see no reason for upper and lower arm and leg bones. Now I'm wondering, for human characters why not put the entire body in one layer -- torso, arms, legs and feet? (I use switch layers for hands.) Has anyone else sorted out their strategy for this?

[Víctor Paredes]

Now I'm wondering, for human characters why not put the entire body in one layer -- torso, arms, legs and feet? (I use switch layers for hands.) Has anyone else sorted out their strategy for this?

I separate layers mostly to animate the layer order and make masks. To have an entire character in one layer it's not too useful (and clean) for the way I rig and animate. Anyway, I think there are some simple characters, like silhouettes, which works fine in one layer.

[Diana Kennedy]

I'm a little stuck with these smartbones.

You install the action at frame zero, right?
Problem is, on that frame, it is'nt possible to adjust the points so the action knows the "target" where to go when bending. It works when I do it in frame 1 but that gives me a first frame I don't necesserly want, you see what I mean?