Hot pixels and intermodulus ones are automatically masked out by pilatus giving them the value of -2. Usually we create manually a mask (at beginning of a run) to take away other 'invalid' pixels, mainly thouse shadowed or perturbed by the beamstop in the beam. However, it can be use for any kind of unwanted features (as a crazy pixel or mysterious cloud suddenly appering in a part of detector area) which can appear any time for various reason. How to proceed to create a new mask:

  • take a 2D image and look on it (ex. Flint);
  • open pyFAI-calib2 (direct gui): 1) fill experiment settings (Pilatus 2M, wavelength,..), 2) Mask: a) intermodulus "-1" b) hot pixels "2" c) beamstop d) flifht tube  and save in /data/bm29/inhouse/Calibration/mask.py once happy (before this save 'old' mask as mask_day_month_year.py);
  • stop (cntrlC) and restart daiquiri (> ./run-server.sh) as opd29 in /users/blissadm/local/bsx3.git in conda environment (>conda activate bsx3);
  • BsxCuBE will ask to re-log;
  • redo a Collect and check if better (on 1D curves).