There’s a rhythm
in that line
We placed the electrodes, recorded a line and checked it for unwanted signals. Now we can look more closely: what speeds are mixed into that line? One complete rise and fall is a cycle.
Move the slider: faster waves pack more cycles into the same second. Hertz (Hz) is simply cycles per second.
And the Greek names?
Delta, theta, alpha, beta and gamma name ranges of speeds. Choose a band to draw its wave below.
Band boundaries vary across studies.2 Page 4 shows how these rhythms are interpreted in different tasks.
Papers behind this page
1. Keil et al. (2022). Recommendations and publication guidelines for studies using frequency domain and time-frequency domain analyses of neural time series. Psychophysiology.
2. Donoghue et al. (2020). Parameterizing neural power spectra into periodic and aperiodic components. Nature Neuroscience, 23, 1655–1665.
All figures are original simulations, not participant recordings.
Make your own mixture
EEG mixes many rhythms with activity that doesn’t repeat regularly. We’ll start with just two waves so you can see how they combine: one at 6 Hz and one at 10 Hz.
A wave’s height is its amplitude, measured in microvolts (µV), tiny units of voltage. Move either slider and watch what happens to the line below.
What went into that line?
Now look at that same mixture by speed. Each bar shows its wave’s power: a taller wave contributes more, so its bar grows. More power doesn’t automatically mean more effort or better brain health.
How do we get from one plot to the other?
Fourier analysis is a way to describe a complicated line as a mixture of simple waves. Here we made the mixture ourselves, so we already know its two ingredients and calculate their powers directly. For these simple waves, doubling the height gives four times the power. Real EEG is a much busier mixture.
With real recordings, researchers often calculate power in short stretches and average the results. How long those stretches are affects what we can see.1
What’s underneath the peak?
Activity that doesn’t repeat in regular waves creates a sloping base in the power plot, called the aperiodic background. If that base rises, power at 10 Hz can rise without a stronger rhythm.2
Lift the background below. The peak moves up, but the bracket showing its height above the background stays the same. This is why researchers distinguish a rhythm from the activity beneath it.
What did this summary leave out?
Power bars don’t show when a rhythm appeared. Page 4 lets you compare a steady rhythm with a brief burst that has the same average.
Why does the window length matter?
Looking at a longer stretch helps separate similar speeds, such as 9 and 10 Hz. Looking at a shorter stretch helps us locate a brief burst in time, but makes its exact speed harder to pin down. 1