When computing the covariance using mne_process_raw and a .cov file it
returns too few samples error.
Then I decided to find the time interval of the raw files and use that for
parameters tmin and tmax in the cov file, and it still returns too few
samples error. Any idea on how I should go about this?