[Mne_analysis] Question about estimating SNR

[Marina Papoutsi]

Dear all,

I have recently started analyzing MEG data using MNE. I am now at the 
stage, where I want to create a movie of my results and I was wondering 
about the value of SNR that I should be using.

I used mne_compute_mne --snronly to get an estimate of the SNR, but I am 
a bit puzzled by the results.
The output is a file called SNR. Based on the MNE_manual the first 
column is the time, the second the SNR+1 and the third the 
regularization factor 1/SNR.
In the file the I got, for one of the conditions, the mean (over time) 
SNR + 1 (the average of the second column in the file) was 0.1  and the 
regularization parameter was constant in time and set to 0.01.

Do these results sound ok to anyone? I am by far not an expert in MEG (I 
have done most of my studies in fMRI), but a 0.1 SNR sounds pretty bad 
to me. On a further note, I am not sure what the relationship between 
the column2 and column 3 is. In the manual it mentions that column 3 is 
1/SNR, but this doesn't seem to quite like that and despite any changes 
in columns2, column 3 is constant to 0.01.

Sorry if my questions are a bit naive. I'm still trying to get my head 
around these things and your help will be greatly appreciated.

Many thanks in advance,
Marina