Hi, I came across invensense demo video and was amazed on how stable the 3D visualization of the object representation in software with all the signal processing. Since I know that the Euler angle (roll, pitch, and yaw) data produce drifts, I believe that the teapot demo is not done by Euler angle. I would like to know how quaternion data is used to visualize the 3D object representation and if quaternion data drifts just like Euler angle. Thank you.