For my developments I use the position of the tracker camera, however it seems that this position is not very accurate and is dependent on the recalibration point.
I have done the following experiment:
Preparation: - trackercam is at a fixed position - The rift can be placed at two positions: P1 and P2. Both are in range of the tracker cam. The orientation is the same, so the rift is not rotated.
experiment 1: 1. place the HMD at position P1 2. start the program which tracks the HMD, no explicit recalibration is called 3. determine tracker position T1 relative to head pose at recalibration (OVRManager.tracker.GetPose(0).position in Unity) 4. determine the hmd position H1 relative to the recalibration point (InputTracking.GetLocalPosition(VRNode.CenterEye) in Unity) 5. The position of the hmd relative to the tracker cam is H1-T1 = position 1
experiment 2: 1. place the HMD at position P2 2. start the program which tracks the HMD, no explicit recalibration is called 3. move the HMD to position P1 3. determine tracker position T2 relative to head pose at recalibration 4. determine the hmd position H2 relative to the recalibration point 5. The position of the hmd relative to the tracker cam is H2 -T2 = position 2
Position 1 and 2 should be the same value, because the tracker cam position and P1 are the same, so the relative position should be the same. However, I see differences of multiple centimeters. The recalibration point is different between the experiments, which may be the cause of this. When I run experiment 1 multiple times I do get the same values. When I run experiment 1 after experiment 2 (thus without touching the HMD), I get different values.
Additional info: - SDK v0.7.0.0 - Unity Personal v5.2.1p3
