| NJ5 said: I think the first two examples can be done with quite coarse positioning, even without WM+. You may have a point with the boxing example though, time will tell. I suspect WM+ is accurate enough for that , but of course you would need two Wii Remotes, the nunchuk is not good enough for that.
|
Maybe you could do the first two with indirect position tracking, but not without WM+. You need it to distinguish between an acceleration and gravity, basically because using the WM+ you know which direction is down in an indipendent manner, and thus you can subtract a vertical 1g acceleration from the numbers registered by the accelerometers in the Wiimote.
The trouble with indirect tracking is that it would probably fail depending on the velocity with which you transition between stances, accumulating errors. As in, the slower your movements the less precisely they would be tracked. I read on a Wii hacker's blog that the Wiimote uses 8 bits for accelerations, with 1g being about 20% of the scale, something like 53 out of 0-255 range. If it scales linearly, this means that the Wiimote will fail to register accelerations of less than about 1/50 of g (about 20 cm/s^2). That's a slow movement, but perfectly plausible in a game action and once it fails to register even for say a second, you'll find your real arm and your virtual one displaced until the next IR calibration happens.
Even more errors will accumulate if the gyroscopes in the WM+ aren't calibrated often enough.
Talking of which, I don't know how calibration is supposed to work in Sony's wands.
