one more example of software hacker trying to do hardware- results are mega overkill :(

aehm… without reading throw all the infos,

but is it not more likely to use the hall sensors in the device for such a application, to get the tilt information of the phone as an absolute data?

Maybe I’m completely wrong but the acceleration data as some kind of relative data would only tell me whether the phone is just moving or not. If I carefully release the phone (with very little acceleration), I could go back to the initial state without change anything….

@T:
You forgot the gravity.
With the acceleration sensor you get a vector facing down, so if the phone is not in motion, you know the attitude of the phone.
But moving the phone upside down downwards with 2g would produce the same sensor output as lying it on an even ground -> 1g.

the accelerometer measures strain (deflection) of a very small cantilever that is affected by gravity. The sensor reports X,Y or X,Y,Z orientation by using gravity, so it isn’t the same as the ones we made in physics class with a rubber band to ride on the roller coaster.

I don’t see the video········