a method of animating natural movements

since softbodies are becoming more workable, i felt this was worth mentioning. this is a simple way to get natural movements using an IK solver and a copy loc constraint, and the record mouse function.
1–make a bone chainn with an IK solver. if you don’t know how consult the documentaion or gabios animation thread.
2–add an empty
3–select the IK solver in pose mode, and click on the button with three arrows on it to get to the constraint menu. add a copy loc constraint, and in the field that says ‘object’, type ‘Empty’. ( without the quotes )
4–now select the empty and open your IPO window. click, and then shift click ( to select them both at the same time ) on the ‘loc X’ and ‘loc Y’ ipo buttons,then… press R for ‘record mouse’… choose ‘still’ from the popup window… press and release ctrl to start and stop recording mouse movements. when your done, left click with your mouse to bake those movements into an IPO.
5–press alt A, and watch the fun.

bump

This looks like an interesting technique … I’ve always been afraid of that R key.

i’ve been dropping hints to the coders that it could use some attention. the current functionality is somewhat limited by a few things:
–can’t play audio strips in the sequencer while using R.
–object is linked to mouse before youu have a chance to position the mouse over the object
–the default scale for movement is too tight, and should have a variable setting.
–no way of setting the recording rate, which currently adds a keyframe for every frame, which is overkill in most cases.

Thanks for the good rundown, Modron. I’ll be trying this, as well as countless other animation methods when I go to animate Yoda (or homer, whichever I get to first).

i should add that if you want to do something complex, parent the empty to another empty, so you can adjust its initial location after you do the recording. it tends to be off a few units some of the time.

Is it impossible to just grab the actual bone in pose mode and animate that way? in my tests it seems not to work.