But when I try plotting the PSD of the continuous EEG signals. The power levels are very high and have a weird peak around 25 Hz, as you can see from the picture attached below. Is it recommended to have PSDs around 0 dB?
Thanks for the clarification again!
Since power line frequency in your recording seems to be 50 Hz, the peak at 25 Hz is likely a sub-harmonic of the power line noise.
Power is usually higher for lower frequencies, so the PSD traces going up as we move from larger to smaller frequencies is to be expected.
Regarding the absolute values / magnitude of the power, this strongly depends on your recording device.
In summary I would say, there’s nothing to worry about in the figure above, except maybe that one green channel which is a bit of an outlier in the sub-10-Hz range. I’d take a closer look at that one to ensure it hasn’t captured some artifact.
Best wishes,
Richard
