[Eeglablist] What are the pros and cons of different references for resting-state EEG?

[Deepa Gupta]

Hello David,

In my opinion, for resting-state EEG source localization:

   - Linked mastoids/earlobes: Easy and historically common, but not ideal
   because the reference itself contains some brain activity and so, can
   distort source estimates. - not a good option.
   - Average reference: Good default, especially with 64 to 128+
   well-distributed electrodes. Simple and robust. - If you imagine this
   scenario, you can realize that it is basically accounting for the entire
   head topography as the baseline and hence, I personally find this the best
   option.
   - REST (reference at infinity): Often best theoretically for source
   localization, because it approximates a reference at infinity and accounts
   for volume conduction/head-model information. - I have personally never
   tried this though.
   - CZ or FCZ: Fine as an acquisition reference, but usually re-reference
   before source analysis - also commonly used by many labs.

Conclusively: For source localization, I’d generally choose average
reference rather than linked mastoids. Make sure the EEG and
forward/lead-field model use the same reference.
Regards,
DG


On Wed, Sep 16, 2026 at 4:02 AM Deepa Gupta <deepag1 at umbc.edu> wrote:

>
> Deepa Gupta <https://urldefense.com/v3/__https://www.linkedin.com/in/deepaguptaresearcher/__;!!Mih3wA!BZc_u1hL3D21Z4wA28yjEd_Vq2R6YNMZq5GL3qQwcchmlGFp-GLsZN3jteUIwYqAPfzn0Yd18JnTq2qDGzON$ >
> Former Postdoctoral Research fellow, AI-READI and INC, UCSD | Ph.D., UMBC
>
>
>
> ---------- Forwarded message ---------
> From: Joseph Dien via eeglablist <eeglablist at sccn.ucsd.edu>
> Date: Mon, Sep 14, 2026 at 8:42 AM
> Subject: Re: [Eeglablist] What are the pros and cons of different
> references for resting-state EEG?
> To: <eeglablist at sccn.ucsd.edu>
>
>
> Hi David!
>
>     It doesn't really matter.  For all the methods I am aware of, they
> either rereference internally (i.e., to average reference), or they use
> a method that is insensitive to reference choice, like CSD.  The only
> issue I can think of is REST, where you can't rereference it, so you'd
> need to be using software that allows you to turn off the internal
> rereferencing, for the ones that do that.  You'd also need to make sure
> all the electrodes have the same reference, so including some bipolar
> ones (e.g., left HEOG minus right HEOG) with the common reference ones
> would not work. Also, you'd want to make sure to include the reference
> channel, which would be all zeroes.  Exceptions would be average
> reference (it's fine if it gets rereferenced) or mean mastoids where
> you've thrown away one or both mastoid channels.  The case where you've
> only got one of the mastoid reference channels is fixable as the missing
> one is simply the mirror image and could readily be added back in by
> yourself.
>
> Cheers!
>
> Joe
>
>
> On 9/10/26 17:06, Gilbert, David G via eeglablist wrote:
> > What are the pros and cons of different references for resting-state
> EEG? In the past many people used linked mastoids or earlobes for
> topological analyses, but what for referencing is best for source
> localization?
> > David Gilbert
> > _______________________________________________
> > To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu or visit
> https://sccn.ucsd.edu/mailman/listinfo/eeglablist  .
>
> --
>
> --------------------------------------------------------------------------------
>
> Joseph Dien, PhD
> Senior Research Scientist
> Department of Human Development and Quantitative Methodology
> University of Maryland, College Park
>
> https://urldefense.com/v3/__http://joedien.com__;!!Mih3wA!HbFInXlwiUruOQYUPN4suLXtD_rUiKl4CT9RzBXJRgi016IUyICFKQU7LFz6kk-1QJCWNFsSqdfY_ST1r2s$
>
> _______________________________________________
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu or visit
> https://sccn.ucsd.edu/mailman/listinfo/eeglablist .
>