Optical motion-capture noise floor: Static-marker recordings supporting Jensenius et al. (2012)

Technical deposit supporting Jensenius, Nymoen, Skogstad & Voldsund (2012), A Study of the Noise-Level in Two Infrared Marker-Based Motion Capture Systems (Proc. Sound and Music Computing 2012). It documents how still a static marker appears to a marker-based optical system - the noise floor against which human standstill micromotion (~6.5 mm/s) is read. Contents: the final paper (PDF + LaTeX); Data set #3, the dedicated multi-static-marker rate sweep recorded at 200/100/50/20 Hz (Qualisys); the noise floor for Data sets #1/#2 (static reference marker extracted, non-identifying, from the Sverm standstill recordings) as per-recording metrics plus two example trajectories and the extraction script; and the original MATLAB analysis scripts. Key result: static-marker noise QoM scales with rate (200 Hz -> 2.5, 100 -> 1.5, 50 -> 0.9, 20 -> 0.3 mm/s) and is far worse outside the calibrated volume. The human standstill data are not duplicated here (they are in the separate Sverm collection); this deposit keeps only the non-identifying reference-marker (noise) data. OptiTrack recordings referenced by the paper's figures were not located (figures only).