[Eeglablist] Discrepancy in Power Values

Marco Montalto montaltomarco at onvol.net
Thu Sep 22 09:18:18 PDT 2011


I am getting a discrepancy between power values in individual subjects' power spectra (plotted in STUDY design) and power values for individual subjects obtained from STUDY design using the following code (using channel Fpz as an example):

[ STUDY specdata specfreqs] = std_specplot(STUDY, ALLEEG, 'channels', { 'Fpz' });

I've been told that when precomputing power spectra measures using specmode "fft", numerical values obtained using the above code are power values (10*log(microvolt^2)), and therefore I understand that they should be the same as the values in power spectra since power in the power spectra is (10*log(microvolt^2)).

Does anyone know the reason for this discrepancy?

Thanks and regards,
Marco






More information about the eeglablist mailing list