Many phones get this wrong (and HTML5 implementations too). Android's
documentation on getOrientation from a rotation matrix is also incomplete
and ambiuguous, and there own implementation doesn't seem to follow what
they document. I believe this is a poor attempt to cover there gaffe of
I believe they are computed from the quaternion, which is in turn computed
from the fusion of magnetometer, accelerometer and if available gyroscope
sensors. Lance has pointed to some good reading the the Android SDK docs,
but these are not complete. For example, the documentation does not
SensorManager's getRotationMatrix() and getOrientation() methods are
interesting to look at. You can see their implementation here:
See my blog for code to do the processing:
http://mysticlakesoftware.blogspot.com/
Rud
On Oct 4, 7:57 am, DD daviddiaofri...@gmail.com wrote:
Hi,
It seems that for onSensorChanged(), the orientation readings
(yaw,pitch,roll) are not detected directly, but computed from the
acceleration
Thank you very much, Rud. I'm looking at your blog. One more thing I
want to ask you. On your blog, you have a piece of code:
private class OrientationListner implements SensorEventListener {
...
Currently, I'm still using SensorListener, and I think it's already
deprecated. But when I changed
Hi Rud, I just finished reading all on your blog. For your reply to my
first question, how Android computes the orientation from the
acceleration. I think only one article is sort of related to my
question, which is
13 July 2009
Sensor - Accelerometer Magnetics
But...er...sorry, I can't really
6 matches
Mail list logo
