Troubleshoot subject calibration

This troubleshooting section covers the following aspects of creating and calibrating a subject in Shogun Live:

A-pose booting

• My subject doesn’t pass the A-pose detection stage.
• The system says “Awaiting boot”.

My subject doesn’t pass the A-pose detection stage.

Look in the log for an explanation as to why the A-pose cannot be detected or why it has been rejected.

• Are any markers reported as missing? If yes:
  • Is the correct skeleton template being used?
  • Are all the required markers are attached to the subject?
  • Does the subject calibration volume include all the reconstructions?
• Does the subject also seem to have labeling issues? If yes:
  • If labeling clusters are attached to the subject, have you created them in Shogun? Have you verified that the labeling clusters you have created are selected?
  • Is the subject standing in a valid A-pose? A valid A-pose assumes that the subject:
    • Stands straight
    • Has their feet pointing forward (ie, the feet are not splayed)
    • Has their palms pointing to the floor
    • Poses their arms at an angle of 45 degrees from the trunk
    • Poses their arms in such a way that they lie in the plane of the torso.
• Are there any joints reported as out-of-range? If yes:
  • Is the A-pose valid (see the above criteria)?
• Is the A-pose still rejected after you've completed all the previous checks? If yes:
  • On the Tracking panel, go to the Subject Calibration section and increase the value for A-Pose Joint Range Threshold (default value: 1.25) to 1.75 and check if this solves the issue.
    Note that subjects who cannot stand in an A-pose as well as small subjects usually require the higher threshold. If you don’t need a higher threshold, use the default value.

The system says “Awaiting boot”.

This message means that the A-pose has been detected and successfully sent to the tracker subsystem. However the tracker subsystem is unable to label the subject from the current pre-calibrated skeleton.

• Is the subject standing in a valid A-pose?
• Does the subject calibration volume correctly exclude marker reconstructions that are not related to the subject under consideration?

Subject calibration

• Calibration of the solving skeleton is jittery.
• The result of the solving skeleton calibration looks bad.
• The labeling clusters won’t attach to the subject.
• During subject re-calibration, the labeling clusters were detached from the subject.

Calibration of the solving skeleton is jittery.

During live subject calibration, jittery refinements of the solving skeleton may occur with some subjects.

Jittery refinements are unimportant as long as the subject obtained at the end of the calibration process is correct.

The result of the solving skeleton calibration looks bad.

Capture a new ROM and ask the subject to:

• Spend more time moving the legs, waist and arms first. Make sure that the motion includes all the joint capabilities. For example, lunging is not sufficient to calibrate the legs. Think of side motion as well and, if possible, intermediate poses obtained by rotating the thigh joints.
• Remember that the more time that is spent on these limbs, the more frames the optimization process receives.
• Avoid a ROM where legs are constantly static and only move in a limited way at the end of the ROM.

Still observing issues?

• Report the problem to Vicon Support, with relevant data. In addition to helping to find a solution to your issue, your input also helps to improve solving skeleton calibration in future releases.
• Use Shogun Post’s Auto-Skeleton feature to obtain a new solving skeleton. Issues with the solving skeleton do not affect the labeling process during live subject calibration, so you can directly re-use the labeled ROM from Shogun Live for Auto-Skeleton.

The labeling clusters won’t attach to the subject.

Ensure at least 10 calibration frames are collected. Shogun only collects these frames if the ROM includes some motion.

During subject re-calibration, the labeling clusters were detached from the subject.

This is expected behavior at the start of the re-calibration process. At least 10 calibration frames must be collected before labeling clusters are re-attached to the subject. This is not possible unless the ROM includes some motion.