When debugging and inspecting the variable S_recon in _prep_maxwell_filter/mne.preprocessing.maxwell_filter, it changes with or without a fine calibration file. The reason seems to be that info[‘chs’] has chaged when passed into _update_sensor_geometry function, which runs only when the fine calibration exists.
In my understanding, S_recon is a forward operator that calculates data on the “idealized” channels from once estimated (by incorporating calibration) multipole moments, thus it is irrelevant to the fine calibration and should not be changed with or without it.
But I am not sure something unexpected does happen in _prep_maxwell_filter, or my understanding is just incorrect.
Thank you @larsoner@mscheltienne for your replies and working on the necessary corrections!
I am glad if you could include me as a co-author, but honestly I don’t have much experience with collaborative development using GitHub, although I do have my account.
This is my account page: link. Is this sufficient information?
@motofumi-fushimi If the GitHub account is the link/page you want to associate, then yes.
To add you as contributor, we need to fill this file with your name and a link to either your GitHub page or to a personal website, e.g. like here.