The excitement over Nintendo’s demonstration of the Wii’s motion-sensitive controller at E3 almost overshadowed the public outcry over the console’s new name. From watching Shigeru Miyamoto conduct a virtual symphony to seeing four people (including a lucky contest winner) play a quick pick-up game of doubles tennis, the potential for new and innovative game play from the Wiimote is nearly limitless.
People were so excited over this technology that, from all appearances, Sony decided to quickly jump on the bandwagon and add motion sensitivity to the PS3 controller, giving its poor Warhawk developers a mere 10 days to include the new functionality in time for E3. However, Sony’s system handles only relative motion, such as tilt on the x, y, or z axis, and quick movements in each of these three dimensions. It works great for games like Warhawk (assuming the lag issues can be worked out) where you are basically just tilting around your jet fighter, but it cannot determine absolute positioning for games like Red Steel that use the Wiimote as a tool for pointing at things on the screen.
Now, in a new article, author Dean Takahashi wonders exactly why that is so. He argues that the Wii controller, which contains MEMS (micro electro-mechanical system) devices, can’t simply keep track of motion without requiring the “control strip” placed in front of the player’s television. The MEMS are used in devices like laptop hard drives to sense sudden excessive motion and park the hard disk drive heads so that there is no damage from the impact.
Takahashi wonders why the Wiimote can’t simply use the results from the accelerometers to determine absolute position, after an initial calibration:
“The chip senses direction and acceleration by measuring changes in the electrons within it.” […] “Broadcom makes the wireless Bluetooth radio chip that transmits the positional data to the console in real time. All of the components are inexpensive and they are a lot faster than they used to be. That’s how a big backhand stroke by the player gets translated into a virtual stroke in a tennis game. But if all that is in there, what do they need the optical sensor for. Can’t they just calculate where the controller is pointing based on the XYZ data and motion data alone?”
The answer seems to be that the accuracy of the accelerometers is not good enough to maintain an absolute position without drift. A player might move the controller really rapidly to the right, then slowly back to its starting position, and a small error in the calculation of the acceleration would mean that the game would misinterpret the motion and think the controller is in a different spot. The Wii uses the three LEDs in the control strip to provide a central calibration point that eliminates this problem by recalibrating the controller’s position every so often.
One drawback of this approach is that the Wii remote will only be well calibrated at a certain distance away from the control strip. This appears to be verified by reports from E3 that the controller will seem a little “off” if the player moves too close or far away from the strip. Left, right, up, and down motions are not affected by this, which is what makes the dramatic tennis strokes seen in the E3 demos possible. However, if the player were to get excited by a drop shot and run up close to the TV set, the sensors would be put off.
Until more people have a Wiimote in their hot little hands, it will be difficult to figure out the exact limitations of the system. However, from all reports it seems as if the controller does work as advertised, and that all the components (including the control strip) are necessary. This means that games on the Wii will not be easily ported to systems like the PlayStation 3.