"SoftBone" is this practical?


#1

I’ve an idea of “SoftBone” recently,
a check box under the Bone info window, once the softbone activate, a control handler appear at the center of two bones, and the object skined to the two bones will become “smooth” (not a good word :P), the control handler is to modify the curvature.
the main



imagine we can easily create a cartoon type of arm like this…


#2

Very nice idea. This could have helped Ryan and I greatly with his last project.

Submit the idea to EI… then pray :wink:


#3

I would think a spline to be a realistic solution… but there may be a way to think of a creative solution… like a plug that simulates it?

Cj


#4

Interesting concept. I’ve never seen a bezier influenced bone system like this before. However, ultimately, I don’t think its practical. The deformation of geometry is based off the influence generated by the position of bones when it is skinned to the skeleton. The more control you want over the mesh, the more bones you’re required. Most 3D systems have mechanisms to modify the area of influence around the bones, which intern effect which vertices are displaced/moved. Weight maps further tweak those base algorithms by adjusting the areas of influence. In your example, its really the bezier spline that seems to be influencing the deformation of geometry more, rather than the bones. (Or at least more so). It would be tough finding a way to average between the bones’ influence and the spline’s influence. Which takes priority?

What we ultimately need in EI is a decent Spline IK system. Its similiar to your concept yhloon, but requires more bones. The chain of bones are linked together with an IK handle and the entire skeleton within the Spline IK is driven by the position of the spline itself. (Including the spline’s CVs)

Internal support for splines in EI, other than animation splines, is drastically required. They would go a long way for character animation and could dramatically improve plugins that are dependant on paths to function.

Though I must admit, the bezier idea is cool, just a bit tricky to implement.


#5

I have played with the idea and even managed to do something like this, only as a proof of concept though. But it works.
The trick is put in a chain of bones (say 10 bones) that deform the mesh and control those 10 bones with two control bones. All you’ll need then is a script and some slider that decides how the rotations from those two control bones get distributed over the ten bones: all rotation in the elbow area (sharp corner) or evenly spread out (bendy).
It’s not something you can do in 5 minutes, that’s for sure. And there may be some discrepancies between the position of the control bones and the actual rig. In other words, the hand may move position as you change the “tension” in the arm.


#6

Thanks for all the replys :slight_smile:

Richard…
having a test on a “round arm” recently, I try use bezier deformation but no luck cause me a lot of troble, than use a lot of bones also but I lost the control of the curve… than I think of combine the best of two, just hope someone could work on it…

CJ, Brian…
thanks for the explanation Brian, too bad the concept is not practical, hopefully EI will implement Spline IK in Version 7.

Manu…
Can you post a project file? highly appriciate :applause:

Loon


#7

I could see this. Its kind of like creating a manual spline IK system, but oey, the hassle. Everyone say, SPLINE IK, to EITG… maybe they’ll listen.

:slight_smile:


#8

Here you are. It’s very primitive. At he moment it only works on the Y-axis. Mind you, that’s probably all you need for any limb.
The other problem is, as I said earlier, that the “wrist” moves up and down as you change the tension in the arm. More so at more extreme angles. You probably can compensate for that in the script, but I couldn’t be bothered. This is just a proof of concept.

just “Save As…”:
http://www.funnylittlemen.co.uk/Temp/SplineIK.prj.zip


#9

I suddenly remembered that in my earlier attempts, I tried to work with IK. Using he control bones to place an IK target for the spine, but it was very prone to unexpected twisting of the spine. I still think it’s a valid avenue to explore, I’m sure there’s ways to avoid the twisting by putting enough limits on the joint-links.


#10

I do have an XP spline IK solution… works fairly well… I have been waiting to a new XP with subroutines before releasing it as with a subroutine I can get greater accuracy in the solver. But it is a real spline solution.

http://homepage.mac.com/bergcj/movs/spline_ik.mp4

Cj


#11

Yes CJ, I have been waiting for you to finish this little gem. Xp 4… where are you!!! or is it 5?


#12

Cool. I wouldn’t animate limbs like that, but definitely a tail or so. How responsive is it when you drag the controllers around?


#13

I feel like you would be able to achieve something very close to your curved bone
sample with some creative wieght maps.
That being the case… we probably won’t ever see a specific bone implemented into the software.
Mike Fitz
www.3dartz.com


#14

The feedback is realtime, and very useful… I dont have Brian’s screen capture software, or I would post a quick interactive movie…

Cj


#15

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.