I'm doing an undergraduate project on the Foucault pendulum, basically just a giant spherical pendulum of which I need to somehow track the motion. I was wondering whether it'd be possible to do this using iPi, since at first glance it's meant for bipeds. Also, should this be possible, how accurate would the measurements be? I'm thinking about comparing the actual motion with the predicted motion in 3ds Max.

Thanks

For a pendulum's motion, you could attach a PS Move to the line. This will give you the rotation data which you can apply the anchor point of the 'pendulum' in your 3D animation program.

To be honest though, it would probably be easier to video tape the motion from the side and then simply match move the swing. :)

For capturing a non-human object moving freely through space, like a tossed ball for example, the item will have to be manually match moved in iPi Mocap Studio--the software itself cannot track non-human objects automatically. You can use the video or point cloud for visual reference.

iPi Mocap Studio does feature props tracking but to capture its full 3D motion, it's assumed that the prop is attached to a performer's limb (typically a hand but not necessarily)--the PS Move and Wii Motion Plus devices only record rotation, not translation.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

That's interesting.

Could you give us some hints about how to use IPI to track arbitrary objects?

I don't have the application in front of me right now test it out... but can you create arbitrary nulls for simple hand-tracking an object through a take? Or do you make a tiny figure and move it around by its root?

In PS Eye camera setups, I do wonder if props could be automatically tracked via isolation by color, e.g. a brightly colored tennis ball.

Oh, that's funny...and clever! I never would have thought of that. :)

Thanks for sharing.

G

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

Just curious, are you trying to show the entire path of the Focault pendulum? Wouldn't it be better to simulate this in a 3D program, probably through math expressions rather than dynamics? I'm assuming you want to accurately show a timelapse animation of the entire path.

Or is this just meant to be an example of non-human motion capture?

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

Thanks for the great tips!

Unfortunately these won't work for
my particular case, since the "bob" is merely
a few centimeters in diameter, not to mention cilinder-shaped.

As for the connecting the PS Move, this won't work
in my case either because the pendulum is actually driven
by a magnet incorporated in the bob itself, and it's mostly the
effect of this driving force that is to be captured.

As for the matchmoving, this could actually work! Though the
pendulum actually starts rotating its swing plane so it would be hard
to capture the full motion using just one camera. But this coul easily be
solved using multiple cameras. Since Ipisoft doesn't support non-human motion,
doing the matching manually would be rather painful since the timelapses will be rather long (30 min. or so) to get a good understanding of the actual motion. I'm not sure if I'm allowed to ask this here, but would you know about any software that does allow for this kind of tracking, or will this be included in future upgrades of ipisoft?

Finally, the option of simulating the motion, I was indeed planning on inputting the equations into 3ds Max, and then comparing that motion to the actual recorded motion. I can't, however, accurately simulate the real motion since it depends on small imperfection in the pendulums construction, which will cause all kinds of precessing motion and whatnot, and my programming skills are nowhere near good enough to incorporate such deviations.

Thanks again for the replies!