EEG Rhythms, Spectra & ERP Averaging

Explore alpha attenuation, normalized Welch power, time-frequency resolution, occipital envelopes, and the limits of ERP averaging.

Chapter 7
EEG
Welch PSD
ERP
STFT

Peak frequency

10.0 Hz

Alpha 8–13 Hz

168.43 µV²

Frequency bin width

0.50 Hz

ERP RMSE / predicted SD

4.90 / 5.00 µV

Continuous EEG; alpha attenuation begins at 4 s

EEG

One-sided Welch spectrum

Welch PSD

Illustrative occipital alpha power

−1250.000 to +1250.000 µV² · blue − / red +

Time-frequency power

STFT power, same window normalization as Welch

Shared power scale: -33.84 to 26.16 dB re 1 µV²/Hz

Independent-trial ERP averaging

True ERPMean of 16 trials

Model scope

Continuous synthetic EEG: rhythm sinusoids plus finite random-phase 1/f-like harmonics; alpha amplitude drops at 4 s and with eye opening. Topography is an illustrative occipital amplitude envelope, not interpolated measured electrodes or a volume-conductor solution. ERP noise is independent Gaussian, 20 µV SD; sqrt(N) gain need not hold for correlated noise or latency jitter. No sleep or seizure diagnosis.

Preparing the simulation…

Conceptual bridges & related modules

EMG Recruitment & Fatigue Indicators
companion · Chapter 7
Compare power-density normalization and physiological confounding of spectral features.
Logistic Baseline: Threshold & ROC Explorer
downstream classifier · Chapter 9 · Section 7
Translate synthetic spectral features into discrimination questions without implying a trained EEG classifier.