[Eeglablist] spectra outputs of spectopo and newtimef (ersp) differ

[Jakob Scherer]

hi all,
i've computed ersp's and spectra using the gui > study > precompute
channel measures. (ersp with 'baseline', NaN, 'freqs', [0 45])
now the values (dB?) of those two methods differ quite a bit. why?
here is how i compared the two methods:

- load the *.datersp file
>> plot(freqs,mean(chan15_ersp,2),'r')
>> hold on
- load *.datspec file
>> plot( freqs(freqs<45), chan15(freqs<45))

the "ersp" curve lies about 50 dB(?) above the "spec" curve and
differs in shape as well. shouldn't they represent the same?


thanks for any hints and comments
jakob