Hello,
I had marked some time segments as bad and proceeded to do Head modelling and its corresponding sources. I performed mtmconvol on the reconstructed source time-series. As seen below, the time interval from 42 to 44 is marked as bad. However, when I plotted the power spectra at different time intervals from the multi-taper time frequency plots. I observe that the red curves from the rejected/bad marked time intervals show peaks like the green curves from the good/acceptable time intervals.
From the tutorials and a previous forum exchange, I was told that bad segments won't be used in the further analysis. I am not sure, why am I getting power spectra in the bad time segments. Can you help out understand ?
]) if there one or more bad events. And it is not possible to do further processing in bad data. If the data was set to be good with the Accept trial option, then the bad segment is changed to good (the "bad" string is removed), and the entire data is used. I hope this makes clear how bad segments are used with imported data.