Alpha Waves The Band Everyone Misreads

Alpha has its name for a dull reason. It was first.

Hans Berger, a psychiatrist in Jena, watched the string of a galvanometer twitch in July 1924, wired to the skin over a skull defect in a seventeen-year-old patient. Recordings through an intact head came later, with a better instrument. The rhythm that kept appearing ran at roughly ten cycles a second, was strongest at the back of the head, and grew when his subject shut their eyes. He called it the alpha wave. The faster, smaller rhythm he picked out underneath it became beta. Only those two are his: delta came from Grey Walter in the 1930s, the remaining letters were handed out by other researchers over the following decades, and the alphabet does not run in order of speed. Berger himself sat on the discovery for five years, published in 1929, and was largely ignored until Cambridge physiologists reproduced the rhythm in the 1930s. It is still occasionally called the Berger wave.

He was not looking for the EEG. As a young soldier during a year of military service he was thrown from his horse into the path of a horse-drawn artillery gun and the driver stopped just short of him. The same day, kilometres away, his sister became convinced he was in danger and pressed their father to send a telegram. Berger spent a career trying to find the physical channel that message had travelled down. He never found it. He found the electroencephalogram instead, which is a reasonable consolation prize.

Alpha switches off faster than it comes on.

Close your eyes and a large 8 to 13 Hz rhythm fills the trace over the back of the head within a second or two. Open them and it does not fade, it collapses, into fast low scribble. This is alpha blocking, sometimes the Berger effect, and clinical recordings include eye-opening and eye-closing on purpose. When it fails to appear, or appears on one side only, that failure is the finding.

If the rhythm vanishes the instant you open your eyes, alpha cannot be relaxation. You are no calmer with your eyes shut. What changed is that the visual cortex now has nothing to process.

Closer to the truth: alpha is what a patch of cortex does when it is not being used. Idling, not serenity. And light is not required to interrupt it. Keep the eyes closed and do mental arithmetic instead, and posterior alpha drops anyway.

Which is worth sitting with if you were about to spend an hour on 10 Hz focus audio with your eyes open on a screen.

Alpha is not the absence of processing. It is a way of shutting processing down where it is not wanted.

The evidence for that comes from spatial attention. Ask someone to watch one side of space and ignore the other, and alpha rises over the hemisphere handling the ignored side. It rises before the thing to be ignored appears. That is not a region falling idle. That is a region being turned down deliberately, in advance.

Read that way, alpha belongs to concentration rather than opposing it. Listening hard with your eyes closed involves strong occipital alpha precisely because vision is being pushed out of the way. The rhythm is doing work.

Alpha is also not one thing in one place. A rhythm in the same 8 to 13 Hz band runs over the sensorimotor strip, called the mu rhythm, and it blocks when you move your hand rather than when you open your eyes. Same frequency, different cortex, different trigger.

And no EEG trace is ever one band. Every recording contains delta, theta, alpha, beta and gamma at the same time; what changes is which one dominates where. Being in alpha means alpha is currently the largest thing in the trace over one part of the head, not that the rest stopped. So alpha as a whole-brain condition is not really how the band works. It is a family of local rhythms that happen to run at similar speeds.

Loops between thalamus and cortex are the leading candidate for the pacemaker. Nobody has closed the question.

Peak alpha frequency differs between people and is strikingly stable within a person. Two people in the same room can sit at 9 Hz and 11.5 Hz, and each will still be near their own number years later. It is partly inherited. It rises through childhood and adolescence, plateaus in early adulthood near ten cycles a second, and drifts slowly downward from middle age onward.

Clinical EEG leans on this. The resting alpha rhythm is read as the posterior dominant rhythm and checked for three things: its frequency, whether it is symmetrical across the hemispheres, and whether it reacts to eye opening. A posterior rhythm that has slowed below the normal adult range, or that is clearly lopsided, or that fails to block when the eyes open, is a flag a neurologist will follow up. The rhythm Berger stumbled into became a routine diagnostic baseline.

A 10 Hz preset is a population average, not your number, and you cannot measure your own without an EEG. Treat the band as a range to search rather than a setting to trust.

Laboratory work in the 1960s showed that people given a tone whenever their alpha rose could learn to keep the tone on. Press coverage followed, then products: headsets and desk boxes promising the alpha state, sold on claims about creativity and enlightenment.

The problem is the one above. Alpha rises when you close your eyes and stop attending to things. A device that rewards alpha is substantially rewarding you for sitting still in a quiet room with your eyes shut, which is a fine thing to do and is not evidence that a skill was trained. The grand claims did not survive scrutiny. The plausible core, that EEG feedback can shift measurable activity, became a research field that still exists.

What definitely survived is the phrase. "Alpha state" is marketing language with a real rhythm underneath it.

A binaural beat is not a sound. Play 200 Hz into one ear and 210 Hz into the other and there is no 10 Hz component in the air at all. The pulse you hear is built inside your auditory system out of the difference between the two inputs, which is why binaural beats need headphones. That much is uncontested. The disputed step is the next one: that the perceived beat pulls cortical rhythms toward its own rate.

Fourteen studies were pooled for a 2023 PLOS ONE review by Ingendoh, Posny and Heine, and they came out against that premise more often than for it; the authors declined to treat the matter as settled. None of it is specific to 10 Hz. The review is about the method, not the band, and the binaural beats page carries the breakdown.

Schwarz and Taylor, in Clinical Neurophysiology in 2005, did record a steady-state cortical response to a binaural beat, but only at low stimulus frequencies, and it came out smaller than the response to a monaural beat. A signal in the cortex is not an effect on how anyone thinks, works or feels, and it should not be upgraded into one.

The most useful recent paper varied the settings instead of picking one and hoping. Melnichuk, Cooper and Hawk, in Scientific Reports in 2025, randomised 80 undergraduates across sixteen conditions and analysed 64 of them, crossing beat frequency, carrier tone, audio onset and background noise, with each person also hearing a matched pure tone as their own control. Gamma beats on the 340 Hz carrier beat that control on percent hits, p = 0.002. Separately, the groups with white noise behind the tones did better than the groups without, p = 0.033. Neither arrangement touched the vigilance decrement — performance drifted down across the session regardless. Sixty-four people across sixteen cells is roughly four per cell, so no single cell should carry much weight.

The EEG from the same study is the part worth arguing with. Beta beats raised power at 16 Hz and gamma beats raised it at 40 Hz, so something followed. But the carrier tone made no difference at all to that, and the white noise that helped the behavioural score cut the entrainment down, sharply for gamma. Behaviour and brain signal did not move together.

That was gamma and beta rather than alpha, so the beat frequency does not transfer. What carries is narrower: the carrier tone and the layer underneath were not inert in a study built to test them.

For context, the clearest case of an outside rhythm showing up in cortical recordings is visual. Flickering light at a fixed rate produces a following response in the visual cortex reliably enough that photic stimulation is a standard part of clinical EEG. Sound produces steady cortical responses of its own. Dragging an existing brain rhythm toward a beat is a different and much shakier proposition.

Search the band instead of trusting 10 Hz. Run 9, 10 and 11 Hz on the same kind of task on three different days. You are not hunting for a sensation. You are looking for the one you stop noticing soonest. The free generator on this site will run the comparison in a browser: set one ear to 200 Hz and the other to 209, 210 and 211 in turn.

Keep the carrier low and plain. The beat is the difference between two tones, but the tones themselves are what you hear for the next forty minutes, and anything bright turns tiring. Low hundreds of hertz, not up where a tone starts to bite. For reference, the low carrier in the study above was 340 Hz.

Bury the beat under something broadband, and make sure that something has no words in it. It goes below the soundscape, not over it: you are not trying to hear the beat, and a naked pulsing tone at 10 Hz is unpleasant to work beside. Lyrics, podcasts and narration are the wrong bed, because they compete for the same language machinery your task needs. Rain, room tone, broadband noise, piano with no melody you already know. Worth knowing that in the same study the noise weakened the EEG entrainment even as it helped the behavioural score. The two do not have to move together.

If what you want is the eyes-closed rhythm itself, close your eyes. Ten minutes sitting still with no screen will raise posterior alpha more dependably than anything coming through headphones. Alpha audio while you read and alpha the EEG rhythm are two different projects. Be clear which one you are doing.

Give it one task and thirty to fifty minutes, and expect something small: slightly less pull toward your phone, a slightly easier time staying with a paragraph. Switch tasks halfway through and you have tested the switch, not the audio. Anything more dramatic than that is probably the room, the routine, or the fact that you finally sat down.

Anyone with epilepsy or a seizure disorder should clear entrainment audio with a doctor first, in this band or any other.

Alpha in Arkhitec is a layer you build on rather than a button you press. The band search this page recommends is the reason the preset list matters: 9, 10 and 11 Hz all exist as presets, in binaural, monaural and isochronic form, so running the comparison takes three taps rather than a signal generator. Only the binaural ones need headphones.

The 7-layer mixer is the rest of it. Put an alpha beat on the bottom layer at low volume, then build the environment over it: rain, ocean, thunder, forest, white noise, jazz piano, Tibetan bowls. Each layer has its own volume and stereo placement, and the stack reopens as one saved mix. The sleep timer stops the whole thing when it runs out.

The affirmations library is the other use for the band. It is pre-recorded, 24 categories and 8 voices, and it runs through the same mixer, so the beat stays underneath the voice.