[Max]How to make soft IK rig like Brad Noble's skeleton Rig?

#1

I know about rigging in intermdiate level,but i still dont know how to make soft IK.
I think it’s very crucial for animator.]
Can anybody give me some direction or tutorial?
Thx

#2

check out this thread.

hope it helps.

#3

thanks for fast reply,great help.
it looks a complicated thing, but i will go into deep.
Please help me in the next question when i get confuse.
sori for my bad grammar
cheers…

#4

Hey Polimeno, that thread you’ve linked to is really cool.

Does anyone know if there’s a simple solution as elegant as Brad’s which will allow you to lock the end of the chain to the control object as in this article thus allowing for soft IK on stretchy limbs?

Cheers

Duncs

#5

Does brads rig use this method? I thought he had a custom system?

#6

Hi…i had try those method but sometime it works 100% and sometime just 99% when i try the same method (i do more and more with the same bone length and position).It still get snapping little bit.I don’t know why this happen.
And how to combine this method with stretchy bone.I am using brad noble’s method that he posted here.

#7

Sorry eek, I was referring to the method Brad describes here using a lookat constraint and the limit controller.

#8

Hey guys.
I saw that xsi-blog article a while back and converted the formula to make a max version - which I’ve attached.

This isn’t used in my skeleton rig as it uses a different method for stretching limbs, instead I use the lookat and limit controller method. However, I will be using this formula in other rigs. Soft stretchy IK is very nice to have.
:¬)

#9

nice Brad !

btw,
could you tell me more about this arm rig ?

my main questions are:

  • how did you aproach nodes (reverse, multiplydivide, and so on) ?
  • Lock to Control inside Wrist_control and Elbow_control ?
  • the inherits were parent constraints ?

i´m doing some R&D based on ELASTIK, thinking in pin elbow/knee control and breakable chains too…

ps
could you guys give me some suggestions, please ?

#10

Hey Brad, thanks for sharing - this works really well…

#11

can anyone explain what djlane do in his softIK rig (his attachment file here: http://forums.cgsociety.org/attachment.php?attachmentid=122624)
He assign Float Script Controller in Position controller of softIK limiter,here the script in float script controller:

chainlen = 201.134
rootpos = root.pos
cntpos = cnt.pos
curlen = distance rootpos cntpos
diff = curlen - (chainlen - softdist)
da = chainlen - softdist

please download his file for details.

i think his script is not including some point3 value like [x,y,z], it’s like his float script only assigned in X_position or Y_position
but it works well.
so can anyone explain how this script can read as point3 value??
i hope djlane can answer this.

#12

Hi brad,please re-upload your attachment,max file can’t opened.

#13

essentially one is the upNode for the other poleVector.

#14

Hey guys,

First of all I’d like to say a big thanks to you guys for posting this, it really helped me out, and got some gears going:) This was my go at it, using Brad’s approach with the limit controller.
http://www.dimitrykachkovski.com/Stretchy_SoftIK.avi
(It is in Xvid format)
It is an awesome solution, but due to the fact that the limit controller does in fact allow you to reach the limit, you eventually get the snapping part at the very last moment. The approach by Andy basically makes the IK try to reach its maximum to infinity (thanks to the exponent). This allows for a much better Soft_IK result as it seems. I will post soon my result of the soft IK with Andy’s method. Another thing, Brad, in your code you added an extra variable which was unnecessary:

IKlength = distance start IKtarget
limbRatio = limbLength/origLength as float
softStart = origLength - softDist
stretch = 1.0
if IKlength > softStart then
(
shortd = softdist*(1.0 - exp(-(IKlength - softStart)/softDist)) + softStart
stretch = IKlength/shortd
)
limbLength * stretch

you never used the limbRatio, so it is not needed:D

So once again, big thanks guys, really liked this!

#15

So I tried it with the formula control and I have found a weird issue. Not quite sure why it is so, but it seems that there still is a small snap. Why is it there, I am not quite clear. Maybe someone can have a look and let me know? Also this seems like a nicer approach as it keeps the control object and IK aligned to each other unlike the method with the limit controller, where the IK starts lagging a little from the control object.

#16

what Max version all attachment file in this thread?
i can extract those file but can’t open it.
i use MAX 9 SP2

#17

Max 2k8 in my case:)

#18

Yeah, I think the file I’ve attached is also done in max 2008.

I’ll get around to answering some of the other questions when I get a chance (which should be soon).

:¬)

#19

Polimeno: The setup is detailed more in this thread.
I can’t remember much about how it works. I think it was a lot of blending of position constraints and the lengths of the arm parts were either a set unlocked value or a distance between a point and a control. Like eek said, in IK mode one control is the end effector and the other is the pole vector. In FK, they are aim targets. I haven’t continued with that setup beyond the initial experiment because in the end I didn’t really like animating with it.

Dimich, you’re right about that superfluous variable. The setup is pretty rough and ready.

:¬)

#20

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.