I bring this up again in a new issue as I think reading is not a problem (mne.io.Raw() does it) but we may want to treat raws differently when fifs refer to CTF or Neuromag data. For example for noise compensation of CTF data we probably want to apply the gradient compensation. Why not using a flag for the vendor or type of machine in self or info? In my own class I am using something like this in the reader:
if self.meg['vendor'] == 'CTF' and raw.compensation_grade != 3: raw.apply_gradient_compensation(3)
(motivated by: https://github.com/jdammers/jumeg/issues/139#issuecomment-604936076)
I bring this up again in a new issue as I think reading is not a problem (mne.io.Raw() does it) but we may want to treat raws differently when fifs refer to CTF or Neuromag data. For example for noise compensation of CTF data we probably want to apply the gradient compensation. Why not using a flag for the vendor or type of machine in self or info? In my own class I am using something like this in the reader:
if self.meg['vendor'] == 'CTF' and raw.compensation_grade != 3: raw.apply_gradient_compensation(3)
(motivated by: https://github.com/jdammers/jumeg/issues/139#issuecomment-604936076)