Jump to content


Photo

Group constraint woes

Doesnt seem to function

  • Please log in to reply
15 replies to this topic

#1 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 16 July 2014 - 09:14 AM

I'm trying to constrain a bone to follow a single cp.

 

My attempted process:

 

1. I made the single cp a group and gave it a name. "cp group"

 

2. In a pose(on/off), I gave the bone a "group constraint" and set the target as the "cp group."

 

3. But in an action....when the cp is moved....the bone does not move. It just sits there and laughs at me. :)

 

 

Maybe I'm misunderstanding the group constraint a bit. 

 

Anyone have an idea why that isn't working?

 

 

 



#2 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 16 July 2014 - 09:29 AM

OK...   I see that it does work.... But you have to add key frames and scrub through the keys to see it.  That's

odd because you can't test if it's working by just moving it on a single frame.

 

Has anyone else had this result with the group constraint.



#3 robcat2075

robcat2075

    persistent smarty-pants

  • Hash Fellow
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 24056 posts
  • Robert Holmén
  • *Moderator*
  • Dallas, Texas
  • current
  • Windows
  • Programmer:NO

Posted 16 July 2014 - 09:41 AM

I just tried it and i get the same behavior. 

 

It may be this is something calculated not continually in real time but only once, when a frame is arrived at.

 

Scrub early and scrub often would be my advice here.  :) 


Robert Holmén
------

Got an A:M question? Come to Live Answer Time.   Saturdays, Noon CDT (1700 GMT)

Watch the 2017 "Summer Memories" Image Contest Awards

 

My tutorials All my most beloved tutorials in one convenient location. Except for the ones I've forgotten about.
 
this is only a ... my gallery of A:M tests

87,848 pushed!: the #1 heavy push on Youtube

Big thanks to... Roger (again!), Shelton (it's huge!), NancyGormezano, Roger, cribbidaj, thefreshestever, Tom, Dalemation, Simon Edmondson, thejobe, Rob_T (2 more x), agep (again!), itsjustme, jason1025(+1), dblhelix (+1),markw, Roger (3x!), mouseman (x 2!), Xtaz, agep, Gerry, thefreshestever, dblhelix (twice!), jason1025, Luuk Steitner, PDM, Rob_T and Dhar!


#4 robcat2075

robcat2075

    persistent smarty-pants

  • Hash Fellow
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 24056 posts
  • Robert Holmén
  • *Moderator*
  • Dallas, Texas
  • current
  • Windows
  • Programmer:NO

Posted 16 July 2014 - 10:34 AM

The thing I've found odd about the group constraint is the way the attached bone behaves.  The Rotate manipulator seems to be in some other universe. :unsure:


Robert Holmén
------

Got an A:M question? Come to Live Answer Time.   Saturdays, Noon CDT (1700 GMT)

Watch the 2017 "Summer Memories" Image Contest Awards

 

My tutorials All my most beloved tutorials in one convenient location. Except for the ones I've forgotten about.
 
this is only a ... my gallery of A:M tests

87,848 pushed!: the #1 heavy push on Youtube

Big thanks to... Roger (again!), Shelton (it's huge!), NancyGormezano, Roger, cribbidaj, thefreshestever, Tom, Dalemation, Simon Edmondson, thejobe, Rob_T (2 more x), agep (again!), itsjustme, jason1025(+1), dblhelix (+1),markw, Roger (3x!), mouseman (x 2!), Xtaz, agep, Gerry, thefreshestever, dblhelix (twice!), jason1025, Luuk Steitner, PDM, Rob_T and Dhar!


#5 Rodney

Rodney

    A:M Bot 14309

  • Admin
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 6979 posts
  • Rodney Baker
  • *Admin*
  • Illinois (not Japan)
  • subscriber
  • Windows

Posted 16 July 2014 - 03:08 PM

I played with the Group Constraint awhile back and I seem to recall hitting the space key to refresh the position of objects constrained to a group.

At issue I believe is the fact that unless the geometry location is keyed A:M has no other way to know where that group is.

Unless keyframed, as far as A:M is concerned the group hasn't moved.

 

Hope that makes sense.

 

For what it's is worth... I kept trying to use the Group Constraint in reverse... i.e. constraining a Group to a Bone.

But that isn't how it works.  ;)


"Animation is 90 percent hard work.  The other half is entirely mental!"
See my effort to think about the art of animation at: My Blog
Want to learn A:M? Start TaoA:M

#6 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 16 July 2014 - 04:43 PM

That makes sense. 

 

Rodney.....Do you remember if there is any rotational control? The rotation of the bone seems to stick to one direction.



#7 robcat2075

robcat2075

    persistent smarty-pants

  • Hash Fellow
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 24056 posts
  • Robert Holmén
  • *Moderator*
  • Dallas, Texas
  • current
  • Windows
  • Programmer:NO

Posted 16 July 2014 - 04:47 PM

 

.....Do you remember if there is any rotational control? The rotation of the bone seems to stick to one direction.

 

So far, the only feasible method I've found is to click and drag (left-right) on the X and Y rotation properties and monkey with those until the bone is aimed like I want.  The Z doesn't seem to do anything.


Robert Holmén
------

Got an A:M question? Come to Live Answer Time.   Saturdays, Noon CDT (1700 GMT)

Watch the 2017 "Summer Memories" Image Contest Awards

 

My tutorials All my most beloved tutorials in one convenient location. Except for the ones I've forgotten about.
 
this is only a ... my gallery of A:M tests

87,848 pushed!: the #1 heavy push on Youtube

Big thanks to... Roger (again!), Shelton (it's huge!), NancyGormezano, Roger, cribbidaj, thefreshestever, Tom, Dalemation, Simon Edmondson, thejobe, Rob_T (2 more x), agep (again!), itsjustme, jason1025(+1), dblhelix (+1),markw, Roger (3x!), mouseman (x 2!), Xtaz, agep, Gerry, thefreshestever, dblhelix (twice!), jason1025, Luuk Steitner, PDM, Rob_T and Dhar!


#8 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 16 July 2014 - 04:48 PM

Is it also possible that sense the group itself hasn't been rotated, then Hash doesn't know what the rotation is either? I wonder if all the groups would have to be rotated as well? That sounds way crazy :)



#9 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 16 July 2014 - 04:52 PM

RobC,

 

You got much closer than me. When I tried to spin any of the handles or the cursers to change the axis, it seemed "gimble" locked.... Did I spell that right?



#10 robcat2075

robcat2075

    persistent smarty-pants

  • Hash Fellow
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 24056 posts
  • Robert Holmén
  • *Moderator*
  • Dallas, Texas
  • current
  • Windows
  • Programmer:NO

Posted 16 July 2014 - 06:17 PM

The group constraint translates a bone to a group but as the group turns in space it will impart no further rotation  to the bone because translation isn't about rotation.

 

A group that consists of one CP only will probably be extra unpredictable.

 

If you need some reliable orientation of a bone that is constrained to a point how about this

 

-make a Group A of the CP you are interested in.

-make a Group B of a point adjacent to it.

 

-Group constrain a bone to each

-AimAt constrain the Group A bone to the Group B bone

-Add a bone to be a child of the Group A bone, at its axis

-orient that child bone and use it for whatever you needed instead of its parent.


Robert Holmén
------

Got an A:M question? Come to Live Answer Time.   Saturdays, Noon CDT (1700 GMT)

Watch the 2017 "Summer Memories" Image Contest Awards

 

My tutorials All my most beloved tutorials in one convenient location. Except for the ones I've forgotten about.
 
this is only a ... my gallery of A:M tests

87,848 pushed!: the #1 heavy push on Youtube

Big thanks to... Roger (again!), Shelton (it's huge!), NancyGormezano, Roger, cribbidaj, thefreshestever, Tom, Dalemation, Simon Edmondson, thejobe, Rob_T (2 more x), agep (again!), itsjustme, jason1025(+1), dblhelix (+1),markw, Roger (3x!), mouseman (x 2!), Xtaz, agep, Gerry, thefreshestever, dblhelix (twice!), jason1025, Luuk Steitner, PDM, Rob_T and Dhar!


#11 Rodney

Rodney

    A:M Bot 14309

  • Admin
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 6979 posts
  • Rodney Baker
  • *Admin*
  • Illinois (not Japan)
  • subscriber
  • Windows

Posted 16 July 2014 - 07:34 PM

Rodney.....Do you remember if there is any rotational control? The rotation of the bone seems to stick to one direction.

 

I believe Robert has already answered this for you.

 

I don' t recall the specific problem I was solving when I resorted to using the Group constraint but *I think* in that case I added an addition Orient Like constraint pointed to a Null.

That Null then became my 'world orientation' and all objects/points oriented to the world had that constraint.  When the world rotated all unconstrained objects stayed in place.

This isn't what you want but in some cases might be an approach that works around the fact that the Group Constraint is a translate constraint.

 

As I said though my goal was to get the Group oriented to a Null/Bone/Another Group and that isn't the way the Group constraint works.

For that I just needed to use a Bone in the first place.

 

There are some fleeting thoughts that are gnawing away at me so I'll have to set something up to test.

I recall that something similar was going on with a model and I was missing a pivot... but I thing that was a Prop/.OBJ.

I dimmly recall testing the grouping of multiple subgroups and then moving a subgroup in order to achieve a proper orientation but that didn't supply the answer I needed.

 

More testing...

(If you have a specific setup you'd like tested please share that as it helps to have a specific target to hit)


"Animation is 90 percent hard work.  The other half is entirely mental!"
See my effort to think about the art of animation at: My Blog
Want to learn A:M? Start TaoA:M

#12 Rodney

Rodney

    A:M Bot 14309

  • Admin
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 6979 posts
  • Rodney Baker
  • *Admin*
  • Illinois (not Japan)
  • subscriber
  • Windows

Posted 16 July 2014 - 07:51 PM

It's coming back to me...

 

I was attempting to use a spline to control the movement of a character (like a puppet being controlled by strings).

I don't know where that project file is but here is a proof of concept I just put together than does reorient a bone based on the position of one of two (or both) CPs.

 

The project file consists of a two CP splines with subgroups 1a and 1b. (in the original project I believe it was a three CP spline with Start, Middle and End)

Both subgroups together are named Controller because it is the one that will receive the Group constraint.

 

Moving the subgroups/CPs about in the Chor moves and orients the Bone (that other than the Group constraint targeting the Controller currently isn't attached to anything)

 

A downside of this approach is that I'm not sure how to view CPs and Bones simultaneously.

I'm sure we can but my brain isn't working there either.

 

There was some cool stuff I was approaching but I stopped dabbling for fear of hurting myself in the process.

I do recall that a nifty thing I ran across was how to control things (objects) with one single solitary CP.

This was actually a trick of sorts in that there was/is a second hidden CP.

That hidden CP never moves, never changes orientation, never gets seen and leaves all the fun of animated movement to that other (visible) CP.

 

 

Edit:  I am reminded of one reason I moved on from this R&D and it is that A:M unhides the hidden (orphaned) CPs upon reopening the model (it must assume that we messed up in hiding it because why in the heck would we want a single solitary CP?).  But it works well until we save.  The impetus for this single CP exploration related to playing with some point cloud type operations in A:M. 

Attached Files


"Animation is 90 percent hard work.  The other half is entirely mental!"
See my effort to think about the art of animation at: My Blog
Want to learn A:M? Start TaoA:M

#13 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 18 July 2014 - 05:39 PM

Rodney,

 

That seems way cool. Point Clouds.....wow that really sounds interesting. When I think of point clouds, I think of the newest compositing frontiers going on out there. Where the colors in a frame can be separated on depth levels.  Is that what you were experimenting with? h a fix. ,

 

Robcat,

Got your message with the fix. Robcat was able to add a bone as a child to the "grouped" bone. The new child can be rotated.  I actually added another step to that.  I gave the new child an "orient like" constraint based on the main mover and it worked as well.

 

In essence....the group constrained bones are mostly like transporters/ taxi-cabs.  Unfortunately, it takes a lot more work.  I was hoping to be able to "snap" on exterior models within a choreography.



#14 robcat2075

robcat2075

    persistent smarty-pants

  • Hash Fellow
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 24056 posts
  • Robert Holmén
  • *Moderator*
  • Dallas, Texas
  • current
  • Windows
  • Programmer:NO

Posted 18 July 2014 - 06:04 PM


 

In essence....the group constrained bones are mostly like transporters/ taxi-cabs.  Unfortunately, it takes a lot more work.  I was hoping to be able to "snap" on exterior models within a choreography.

 

Is it possible for you to show me a more specific example of what you are trying to do this with? I can't help but think there's got to be another way?


Robert Holmén
------

Got an A:M question? Come to Live Answer Time.   Saturdays, Noon CDT (1700 GMT)

Watch the 2017 "Summer Memories" Image Contest Awards

 

My tutorials All my most beloved tutorials in one convenient location. Except for the ones I've forgotten about.
 
this is only a ... my gallery of A:M tests

87,848 pushed!: the #1 heavy push on Youtube

Big thanks to... Roger (again!), Shelton (it's huge!), NancyGormezano, Roger, cribbidaj, thefreshestever, Tom, Dalemation, Simon Edmondson, thejobe, Rob_T (2 more x), agep (again!), itsjustme, jason1025(+1), dblhelix (+1),markw, Roger (3x!), mouseman (x 2!), Xtaz, agep, Gerry, thefreshestever, dblhelix (twice!), jason1025, Luuk Steitner, PDM, Rob_T and Dhar!


#15 detbear

detbear

    Chronic Visualist

  • Craftsman/Mentor
  • PipPipPipPipPipPipPip
  • 896 posts
  • - - - -

Posted 29 July 2014 - 08:10 PM

Hey......Rodney and Robcat,

 

I must ask you both to forgive me for not responding to this thread in a timely manner. I had to move on and got

swamped in a different direction. 

 

Thank you both for looking at this with me.  I think in the future, the group constraint has some potential for snapping

things onto a model as an alternate to other methods. But the Group constraint itself has to operate a bit more efficiently

for that to be feesable.



#16 robcat2075

robcat2075

    persistent smarty-pants

  • Hash Fellow
  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • 24056 posts
  • Robert Holmén
  • *Moderator*
  • Dallas, Texas
  • current
  • Windows
  • Programmer:NO

Posted 29 July 2014 - 08:36 PM

In the future. before you embark on this again, lets look at a specific case and see if there is a good way to do it.


Robert Holmén
------

Got an A:M question? Come to Live Answer Time.   Saturdays, Noon CDT (1700 GMT)

Watch the 2017 "Summer Memories" Image Contest Awards

 

My tutorials All my most beloved tutorials in one convenient location. Except for the ones I've forgotten about.
 
this is only a ... my gallery of A:M tests

87,848 pushed!: the #1 heavy push on Youtube

Big thanks to... Roger (again!), Shelton (it's huge!), NancyGormezano, Roger, cribbidaj, thefreshestever, Tom, Dalemation, Simon Edmondson, thejobe, Rob_T (2 more x), agep (again!), itsjustme, jason1025(+1), dblhelix (+1),markw, Roger (3x!), mouseman (x 2!), Xtaz, agep, Gerry, thefreshestever, dblhelix (twice!), jason1025, Luuk Steitner, PDM, Rob_T and Dhar!





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users