Hello MNE users,
Yesterday I was asking about the noise covariance matrix for empty room in order to compute single trial analysis at the source level with MNE.
My specific questions were answers and thanks for this
Unfortunately I forgot to mention that we are combining MEG & EEG for the source files
The problem that I have now is the following:
If I compute the noise covariance matrix with an empty room, obviously I am missing the covariance matrix for the EEG.
Then when I want to compute the inverse operator (mne_inverse_operator) with this empty room covariance matrix, MNE detect that in the raw fif files there are more channels than in the covariance matrix files (error message: all desired channels not found in the covariance matrix, at least missing EEG001)
Does this mean that for combined MEG&EEG single trial source files we have to use "a normal" covariance matrix file and not an empty room one? I meant we should compute it in the same way as the one obtain for an evoked data?
Thanks in advance for your help
Elisabeth