Time-frequency analysis / Averaging

Dear Brainstorm Community,
Assuming that I am interested in the amplitude and phase of a particular frequency in the source space, time-frequency analysis is needed. There are two methods of operation:
(1) Source analysis of a trail. Then the source signal is analyzed in time-frequency domain to find the amplitude and phase of the frequency of interest. A number of trials perform the above operations, and ultimately obtain a number of amplitude and phase values. Finally, these amplitudes and phases are averaged to get the average frequency and phase.

(2) Average trail on sensor directly, such as getting ERP signal. Then the source of ERP is analyzed. Time-frequency analysis is performed on the average source to obtain the amplitude and phase of the frequency of interest.

From the Brain Strom Tutorial, we can see that the first calculation method is the best, but it will bring a lot of calculation. Is the error caused by calculating in the second way tolerable?

Looking forwoard to replies.
Ning

Is the error caused by calculating in the second way tolerable?

No it is not tolerable, you don't estimate the same components in the signals.
This is discussed in the articles listed in reference of the time-frequecy tutorial (mostly Bertrand 2000 and Pantazis 2005 I think):
https://neuroimage.usc.edu/brainstorm/Tutorials/TimeFrequency#Additional_documentation

Thanks very much for your kindly reply.