[Eeglablist] PSD artifact in merged EEG data

[Cedric Cannard]

Dear João,

I would suspect a boundary effect caused by baseline shifts or signal discontinuity between EEG_run1 and EEG_run2. Applying DC correction or filtering before merging should resolve the issue in most cases. If not, consider segment-wise PSD estimation rather than merging raw data.

Hope this helps,

Cedric Cannard



On Monday, 7 July 2025 at 11:00, João Paulo Viegas Amaro via eeglablist <eeglablist at sccn.ucsd.edu> wrote:

> Hello all,
> 
> I’m encountering an issue when merging EEG data from two different runs.
> The power spectral density (PSD) plot of the merged data consistently shows
> an unusual artifact — a sort of “bouncy” pattern — which does not appear
> when plotting each run individually. I have observed this in most of my 20
> subjects, and the artifact is almost always limited to the merged data.
> However, in one case, a similar pattern did appear in a single run on its
> own.
> 
> I’ve attached an example from one subject (shared in a Google Drive link),
> which includes the code, PSD plots for each individual run, as well as the
> merged data, for comparison.
> 
> The data were recorded using EGI’s 256-channel EEG system at a sampling
> rate of 500 Hz. The plots were generated immediately after loading the
> data, with no preprocessing or other modifications applied beforehand.
> 
> Has anyone come across this type of artifact or have any idea what might be
> causing it?
> 
> Link:
> https://urldefense.com/v3/__https://drive.google.com/drive/folders/1NWPBNA03aeL0EAheAWbmay41elEqnelq?usp=sharing__;!!Mih3wA!HzTxfZi2NwXU5fzmy2gQR2Lwk8rEyvzE6vm6VZX0Gl4YMmhgA6H-ZPjgMGGQPdN33T4tCVU0b0gg5uRbLQNev2dO1QGnrQ0$
> 
> Thank you in advance for your help!
> 
> Best regards,
> 
> João Amaro
> _______________________________________________
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu or visit https://sccn.ucsd.edu/mailman/listinfo/eeglablist .