Huge Jumps in Euler Data

[DomSte]

I used 8 Flex13 cameras in order to track a musical instrument's position and orientation. The rigidbody located at the instrument consists of 5 7/16″ diameter hard markers.
I did several measurements, all under the same conditions. Most of the results look plausible, however, in some takes, there are huge gaps or jumps in the euler angles.
I imported the CSV-output of optitrack into matlab and made a plot of the angles timeline (see attachment) to show my problem. I also checked the CSV- data to make sure that there is no fault in my matlab script.
Is there any explanation, why, for example the pitch value jumps from 144.7 to -46.5 from one frame to the other (with only 1/120 sec in between)?!
The instrument was definetly not moved that fast.

Thanks in advance