[Eeglablist] Which is the best way to measure the "alpha" oscillation?

[Евгений Машеров]

A double peak is quite common in the alpha rhythm. It can be generated by two mechanisms: either two independent sources of different frequencies (v and w), or a single source modulated by low-frequency oscillations.
sin(vt)+sin(wt)=2sin((v+w) t / 2)cos(v-w) t / 2)
Deciding between these two hypotheses based on a single recording is fundamentally impossible, but if the two independent sources are located in different regions, the problem is, in principle, solvable. Unfortunately, I was unable to do this—the two-source model requires at least twelve parameters (if we assume a single source is dipole, then these are three coordinates and three dipole components for each; source frequencies are also parameters, although they can be estimated separately). A fairly strong manifestation of occipital alpha activity is observed in seven leads of the 10-20% system (O2, O1, P4, P3, Pz, T6, T5). In more distant leads, alpha activity of occipital origin is greatly weakened and masked by the intrinsic activity of these regions.
At that time, we did not have a high-density EEG system, so perhaps it makes sense to revisit this issue.

> Dear all,
> 
> I appreciate some good ideas on identifying alpha frequency peaks. I think
> one interesting part of alpha oscillation is when we start statistical
> analysis across different groups. Depending on environmental or individual
> differences, some subjects show two distinctive alpha frequency peaks,
> while others might demonstrate only one distinctive peak on power spectrum
> (of course there are many cases in which peaks are non-identifiable).
> 
> In this sense, it's not easy to make an a priori decision on whether such
> spectral differences are derived from individual differences on EEG or
> significant group features. Maybe implemtation of different pipelines
> altogether is a rough solution.
> 
> Best Regards,
> Jinwon
> On Thu, Apr 30, 2026 at 9:53 PM Makoto Miyakoshi via eeglablist <
> eeglablist at sccn.ucsd.edu> wrote:
>