Jump to content
Hash, Inc. Forums
Sign in to follow this  
fraizer

Theo Jansen - Strandbeest - Rigging the Leg Mechanism?

Recommended Posts

Background:

Theo Jansen designs and builds creatures he calls Standbeests; they are quite large, wonderfully whimsical, and often wind-driven.  Here’s a link to his website:
 
www.strandbeest.com
 
A few days ago, I decided to model a version of his Rhinoceros, not a particularly difficult task.  Then - before building the model - I fell down the rabbit hole of trying to mimic the leg mechanism that drives the Beest the way it actually works.  Not necessary to make a model animate, as a simple set of bones with basic key framing could do that job, but an interesting problem.  I have been stuck in the rabbit hole for days.  Rigging and constraints are not really my strong suit…
 
The mechanism is deceptively simple; the leg is driven by a crankshaft offset from a pivot point about which the leg rotates.  The motion of the leg is controlled from two directions with linkages fixing the range and direction of the motion.  It took Jansen a while to determine the correct lengths for all the relationships, but once calculated, building a working model was straightforward.
 
Here is a link to a simple wooden mock-up (GIF) with a hand-crank drive.  Also on this page is a video of Jansen pulling his full-size Rhinoceros (as well as many other Beests):
 
 
The reason I cannot get out of the rabbit hole is I have failed to find a bone structure / hierarchy and set of constraints that will drive the leg mechanism and not come apart.  Not even close.  I suspect I am missing something obvious.
 
I did a forum search and found no references to Jansen or his Beests.  I thought that strange as they are so interesting…
 
Here is my simple leg model, without bones (except M, the crank); also the PNG rotoscope file (from Jansen’s site).  The line segments are as close as they can be to Jansen’s holy numbers, in cm.
 
Thoughts?
 
Joseph Fraizer
www.iconmine.com

leg-mechanism-test08-deboned.mdl

509px-Strandbeest_Leg_Proportions.svg.png

Share this post


Link to post
Share on other sites

That is an interesting challenge! I recall seeing videos of those contraptions on the beach a few years ago.

Share this post


Link to post
Share on other sites

With a little background help from Robcat, I have made an action that almost works.  The mechanism moves as it should except for the C-bone which does not do its job.  The action is 2 seconds, 30FPS; by using a key framing cheat at 0:15, 1:00, and 1:15 -- moving the C-bone and all other bones into correct positions -- I get very close to the characteristic gait of the Beest.  Good enough for animation purposes.  But I do not think this approach is the real solution to rigging the leg.

Here is a new version of the model, boned.  And the action I have developed.

 

leg-mechanism-test09.mdl

leg-motion-test09-working.act

Share this post


Link to post
Share on other sites

Very cool.  It works much better than my attempt.  I am still studying your bone, null, and constraint set-up.  Cannot say I fully understand it just yet.  Thank you.

Share this post


Link to post
Share on other sites

I myself do not understand how this works. We need to work together to make a more universal installation for the study of the most optimal solution. Thank.

Share this post


Link to post
Share on other sites

Hi Joseph,

I did a walk thru of this at Live Answer time thinking I could just post the screencam of that for you, but my screencam system seemed not to work and still isn't working when i tried to do it again.

I'll try to get back to you on this.

Share this post


Link to post
Share on other sites

Hi Joseph,

Here is a video walking through the set up. My solution is similar to Serge's

The audio sync is bad, it will wander in and out of sync but i think you can follow it.

 

Share this post


Link to post
Share on other sites

How exactly the mechanism works for you, Holman! It is wonderful!

  • Like 1

Share this post


Link to post
Share on other sites

Here is a zip with the PRJ with the leg both the undone (01) and done (06) form if you want to try it as in the video.

BeastLegPRJ.zip

  • Thanks 1

Share this post


Link to post
Share on other sites

Robert,

A brilliant solution and excellent explanation.  For me, the key concept that I would never have arrived at is making bone F a child of bone B, without any intervening/connecting linkages. And your model and action are far simpler than my almost-works attempt.  Simpler is always better in A:M world.

Thank you.

Joseph

  • Like 1

Share this post


Link to post
Share on other sites

Here is a movie of my Rhino-Beest model (so far).  Looking at the video of Jansen's Rhinoceros, it appears that the six leg pairs (front + rear = pair) each have separate cycles.  I decided to simplify the task of boning/constraining the model/action by making 3 pairs of bones/constraints and assigning CPs to leg bones so that the legs move symmetrically on either side of the model centerline.

The action is driven by bones M1 through M6 which rotate around a common center and each move one of the leg mechanisms plus its respective CP doppelgänger.  The walk cycle that seems to best resemble Jansen's Rhinoceros has pair M1-M2 (front-rear) set at 40 degrees apart, then an 80 degree gap followed by M3-M4, etc.

Thanks again to Robert and Serge.

  • Thanks 1

Share this post


Link to post
Share on other sites

Looks like you'll need a different video format. That one doesn't play. Try MP4

Share this post


Link to post
Share on other sites

It is MP4 in a MOV wrapper.  Is this a Mac vs. Windows thing?  The video plays for me...from the Hash website...

Share this post


Link to post
Share on other sites

I suspect most people will not have MOV available anymore since Quicktime has been declared a security problem. It's off in my browser.

Try just changing the extension to MP4.

Share this post


Link to post
Share on other sites

No, i don't see it yet.  😯 Does anyone else see it?

What was the actual video compressor used?

The MP4 I posted previously was h.264

Share this post


Link to post
Share on other sites

TGA sequence was exported from Quicktime 7Pro compressed as MPEG-4 video.  "Prepare for internet streaming" set to Fast Start.  Default extension is MOV.  I changed the extension to MP4.  Does play for me.

Share this post


Link to post
Share on other sites

Here's what happens for me. I'm not sure what is wrong...

frazierpost.JPG

Share this post


Link to post
Share on other sites

I am puzzled too.  Usually these things "just work."  I cannot effectively test other formats, etc., because everything works for me in my browser (Safari).

Share this post


Link to post
Share on other sites

You could upload it to YouTube and put the Youtube link here.

Share this post


Link to post
Share on other sites

YouTube it is.  Here is the link:

https://youtu.be/yLnrZywVnbA

Sorry about the misfires...

If anyone can provide a Mac-based workflow for reliably producing videos that will play in the forum -- much appreciated.

  • Like 1

Share this post


Link to post
Share on other sites

Now I see it!  It's a beest!

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×