Adequate amount/proportion of data to be marked as bad in find_bad_channels_maxwell()

Hi all,

I am using the mne.preprocessing.find_bad_channels_maxwell() function to automatically detect bad channels in considerably long (~2 h) MEG datasets.

The default duration=5.0 and min_count=5 parameters seem too stringent for such a long recording, so I would like to know whether there are some guidelines (just rule of thumb is fine too :slight_smile: ) on which amount/proportion of the signal in a channels should be detected as bad for it to be considered bad.

Thanks a lot!