Hi, I am doing some research on different motion capture systems in the gaming/animation industry and was wondering if any of you have had any bought, used, or are familar with any of the mo-cap technology out there. I just want to hear some feedback about the different brands that you know about and what you think about them. Also, what kind of technology (ie: optical, inertial, magnetic) is the best for the animation industry and what are some of the advantages or disadvantages between the different systems and technologies? If you have knowledge or experience with mo-cap I would greatly appreciate hearing your input. Thank you

It depends what the application is. I've used a lot of Ascension magnetic mocap stuff, and it was very good, for both our medical and entertainment needs. Plus they provide a full SDK, allowing you to write your own applications, which was very important for our needs.

I know my old place of work uses an Ascension system for both film and game mocap, and are doing very well with it.

I worked on one of the first performance animation systems supporting magnetic, optical and exoskeleton rigs. (Several years ago, but this information may at least get you started.) Which is best depends on your needs.

Magnetic systems suffer distortion from ambient ferrous metal in the capture environment (for example, rebar in the concrete floor, AC wiring in the walls and ceiling, that steel file cabinet you're using to elevate the emitter a couple of feet off the floor; and forget about using a real golf club or sword as a prop). AC systems (Polhemus) have different distortion/interference characteristics from DC systems (Ascension) and require different filtering to get good results. The sensor suits worn by the performers can be uncomfortable and awkward, particularly if tethered rather than wireless. On the plus side, they output 6DOF tuples that are easily mapped to your kinematic hierarchy, so it's easy to preview the capture session on-the-fly.

Optical systems have to deal with tracking occluded reflectors. Optical systems triangulate 2D camera images to 3DOF positions, which must be processed to map to a kinematic hierarchy, and older systems at least couldn't really do that on-the-fly. A reflector may also be occluded with respect to some but not all of the cameras, or with respect to all of them, in which case the reflector is "lost" and the data will need alternate tracking. I think Kaydara handles all of this. The reflector suits are less awkward for the performers than magnetic sensor suits.

Exoskeletons (e.g. Gypsy) don't suffer from electromagnetic interference or distortions, nor from line-of-sight occlusion. And the mapping to your kinematic hierarchy is trivial. But you can't exactly roll around on the floor or breakdance while wearing one, so not all motion is suitable for such devices.

Be aware of constraints on the size of the capture area. And be aware of capture frame rates as well, which often depend on the number of sensors/reflectors in use.

I personally think of optical as owning the general-purpose future, magnetic and exoskeleton for special niche purposes.