I've had trouble animating characters using 3D applications. It is too slow and difficult. I've programmed my own 3D animator that uses only the keyboard to change key angles. The results are much better, yet it draws some considerable time.


I'm reaching the conclusion animating is hard because nor the keyboard nor the mouse are ideal input sources for that kind of thing. I guess the best would be a physical doll one could manipulate. Is anyone aware of such a product?

  • \$\begingroup\$ Physically animating is usually done through inverse kinematics. I'm not aware of specific products, but it is a pretty common thing to do. I'd imagine quite a few popular physics engines have such features, though I'm not sure. The idea to set a few keyframes for your animation and solve for forces to interpolate between frames. Jacobian transpose is the badass implementation. \$\endgroup\$
    – RandyGaul
    Aug 11, 2013 at 1:32
  • 1
    \$\begingroup\$ Motion capture via multiple cameras is the more common approach to this problem. \$\endgroup\$
    – House
    Aug 11, 2013 at 1:39
  • \$\begingroup\$ @Byte56 Hm but how could I setup this? Also problem is I can't make all poses I'd like my characters to... \$\endgroup\$
    – MaiaVictor
    Aug 11, 2013 at 1:48
  • \$\begingroup\$ Oh did you mean like real-life physical? Sorry I thought you meant physically modeled in software :( \$\endgroup\$
    – RandyGaul
    Aug 11, 2013 at 1:51
  • \$\begingroup\$ @RandyGaul Yes, no probs. \$\endgroup\$
    – MaiaVictor
    Aug 11, 2013 at 1:52

1 Answer 1


There is actually a doll for exactly this purpose:

enter image description here

It appears to be under development by SoftEther Corporation.

However, I imagine such a solution might be overkill.

An alternative would be to use a GI Joe for your modeling. Record video of your GI Joe manipulations from the front, side and top. Using these synced videos, you can go frame by frame switching between a front, side and top view of your 3D model (or you may be able to do it with two cameras). For each frame align your model to match the GI Joe in the recorded video for that frame. This is similar to rotoscoping for 2D, but you're doing it on each axis to expand it to 3D.

Another alternative would be to use a 3D camera like the Kinect. enter image description here Remember that you can always elaborate on the animations when you're creating them. You don't need to do a backflip in front of your Kinect to have your character do one. The accuracy may not be great, but it'll give you a good baseline to work off. (However, if you do attempt the backflip, you'll have some sweet 3D video of it.)


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .