Trouble referencing misc channels type of eeg data

Hey. I’m having a trouble when referrencing ‘LM’ and ‘RM’ channels set as ‘misc’ channels. The filtered frequencies of my eeg data emerge as unfiltered when refferencing to mean of LM and RM (misc type) . Is it supposed to happen? The thing is that when I don’t set these channels as ‘misc’ the problem is solved, but I could not configure to 1020 channels positions at first.

  • MNE-Python version: 0.23.0
  • operating system: ubuntu 20.04 LTS

# Referencing 'LM' and 'RM' set as 'misc' types
raw = mne.io.read_raw_cnt(eeg_data_raw_file) # please, use any example you have of EEG data

raw.set_channel_types({'LM': 'misc'})  # left
raw.set_channel_types({'RM': 'misc'})  # right

raw.load_data()
raw.filter(1., 20, fir_design='firwin2')
raw_custom, _ = mne.set_eeg_reference(raw, ['RM','LM'])
raw_custom.plot_psd()
raw.plot_psd()

####
# Without setting RM and LM as misc channels
raw = mne.io.read_raw_cnt(eeg_data_raw_file) # please, use any example you have of EEG data

raw.load_data()
raw.filter(1., 20, fir_design='firwin2')
raw_custom, _ = mne.set_eeg_reference(raw, ['RM','LM'])
raw_custom.plot_psd()
raw.plot_psd()

hi,

raw.filter applied filtering to what mne considers data channels. Misc are not
considered data channels. To achieve what you want you can just pass picks to
raw.filter

HTH
Alex

Oh, thank you so much Alex!
Raul