Recording positions of bones

Hi there,

Last year I’ve implemented in Blender a mechanism like the one invented by Theo Jansen some time ago (plase Google the issue for more info). I strictly followed a tutorial explaning how to connect the bones, which is a parent of which, how all they should be connected to the drive, etc. and it works :wink:

I’ve been recenly interested to record the end point (the foot) of the mechanism for further developments and I made a short script for that purpose. Plase find all in this .blend file…

I have some questions and dont have the answers:

  1. Notwithstanding that mechanism is made under the tutorial by an expert, there is a point which looks “unatached”… I wonder why… and how to avoid this…

  2. I’ve made the script to move the mechanism 3 full cycles. In theory, all three should be identical. I noticed that there is some visual deviation while increasing and/or decreasing poses (frames) manually. Which means the IK solver works not quite good… Why? Due to that reason I made the script running 3 times and not just once and then taking only the second set of coordinates recorded. You can all see that the script records the coodrinates of 1 point from the pointer object at lower part of the mechanism and that the pointer object is a child of the lower bone, so it is connected as it should be. Also why there is a clitch - look at the trajectory recorded (I unrestand that this can be eliminated by cutting 1 of recorded coordinates from the list but… why it is so recorded?) ???

The issue is that as soon as there are deviations between the visual and recorded path, one cannot rely on calculating positions beforehand, right?

Changing lenght of the bones will result in interesting resulting curves so I am interested to experiment with this :spin:

I know that the Drive should be a circle but any rotating part will be ok so pls dont criticize the square shape.

Discussion + piece of advice on how to do it differently WELCOME… :wink: