Forum Discussion
DK2 Camera only tracking
I'm wondering, is DK2 capable of using camera-only tracking, without IMU?
I've tried not to request it by not supplying ovrTrackingCap_Orientation | ovrTrackingCap_MagYawCorrection to ovrHmd_ConfigureTracking(), but it still obviously uses inertial measurements.
The reason for this request is that I'm planning to use the HMD on a motion platform. A motion system does a lot of tricks to fake human sensations, but this also confuses the DK2 IMU. (Say, gravity doesn't always point downwards; some accelerations do not correspond to head movements, etc.)
A natural approach is to use non-inertial measurements, such as the camera on DK2. But can OVR algorithms work in this configuration? I understand it must be difficult, given much slower refresh rate and absence of measurable derivatives, but has anyone tried it? Or perhaps planned for future work?
A work-around is possible where I would measure acceleration/velocity of the motion platform, and then DK2 would need to subtract them from its own measurements. But is this possible? That is, get into the inner loop of the device and make such adjustments? If the inner loop is 1000 Hz as claimed, this must be tricky to do, but I guess Oculus would be interested in some solution that would allow the device to be used in non-inertial systems... They expand the VR effect enormously!
I've tried not to request it by not supplying ovrTrackingCap_Orientation | ovrTrackingCap_MagYawCorrection to ovrHmd_ConfigureTracking(), but it still obviously uses inertial measurements.
The reason for this request is that I'm planning to use the HMD on a motion platform. A motion system does a lot of tricks to fake human sensations, but this also confuses the DK2 IMU. (Say, gravity doesn't always point downwards; some accelerations do not correspond to head movements, etc.)
A natural approach is to use non-inertial measurements, such as the camera on DK2. But can OVR algorithms work in this configuration? I understand it must be difficult, given much slower refresh rate and absence of measurable derivatives, but has anyone tried it? Or perhaps planned for future work?
A work-around is possible where I would measure acceleration/velocity of the motion platform, and then DK2 would need to subtract them from its own measurements. But is this possible? That is, get into the inner loop of the device and make such adjustments? If the inner loop is 1000 Hz as claimed, this must be tricky to do, but I guess Oculus would be interested in some solution that would allow the device to be used in non-inertial systems... They expand the VR effect enormously!
No RepliesBe the first to reply