When animating something like a robot or a vehicle that has clear mechanical hinges or wheels, should one use typical rigging and bone animation methods, or just make the animated regions separate objects to manipulate over multiple keyframes?

For example, I'm making a fictionalized/lower-poly space shuttle in Blender. The wing flaps, rudder, and wheels will be animated. I have made all the parts that I want to animate (like the landing gear and the wing flaps) as separate objects from the main body.

Should I keep these as separate objects and simply rotate them over the desired keyframes? Or should I attach them to the main object, add rigging (with 100% weight to the animated parts and 0% to the main body) and animate the bones over the desired keyframes?

Same thing for robots. Should a (low poly) Iron Giant type robot be animated by manipulating separate objects, or by traditional rigging? What about a simpler wheeled robot?

My goal is to import it into Unity, if that changes anything.

  • $\begingroup$ This is opinion based as it is written, the site rules discourage this because it tends to start fights between people who have a favorite method. You can avoid this by re-wording the question to ask what are some of the pros and cons to using either method. $\endgroup$ – MarcClintDion Nov 23 '14 at 3:57
  • $\begingroup$ For instance, many separate objects will have a GPU driver penalty because many draw calls have to be made for every model. On the other hand, using rigs/weight maps can bloat the vertex shader and cause inefficiency there instead. Separate draw calls for every part of the model does however allow you to use different textures for every part which allows for much higher overall resolution without using huge textures which can stall the pipeline. Of course having only one object with only a couple textures is much easier to manage. There are likely many things that people can add to this list $\endgroup$ – MarcClintDion Nov 23 '14 at 4:01
  • $\begingroup$ @MarkClintDion Thanks for the info. I reworded the question. Basically, will modelling with or without separate objects give more versatility to the animation (particularly after exporting from blender into another application)? As far as the GPU penalty you mentioned, how severe would it be if I had, say, 6 objects, with one having 900 polys and the other five having less than 20? $\endgroup$ – angrycactus Nov 23 '14 at 5:14
  • $\begingroup$ Even a 4th gen ipod touch could handle that with no issue. The thing about performance issues like this is that it's very sensitive to the specific video card being used. About 10 years ago, an ATI programmers document was recommending that VBO's(model data) should be about 1800 vertices. I think that's still good advice if you want to target mobile devices. I've read that some modern high-end games running on modern desktop GPU's go as high as 8000-9000 vert's per model now. There is no rule to fit all the possible situations, it's just something to keep in mind and test often. $\endgroup$ – MarcClintDion Nov 23 '14 at 5:51
  • $\begingroup$ You could armature bones to control the subobjects instead of doing it manually, as a third option. $\endgroup$ – StarWeaver Nov 25 '14 at 21:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.