Alpha Waves: What They Are and How to Listen

Alpha waves are the brain's rhythm of 8 to 13 cycles a second, strongest at the back of the head while you rest awake with eyes closed. Across 109 volunteers' recordings, closing the eyes multiplied alpha power by a median of 3.5. An alpha binaural beat, like one gliding from 12 to 9 Hz, is a listening choice: studies of whether the brain follows it are small and mixed.

Ready to play Falling Alpha Session 12 to 9 Hz

Open the 12 to 9 Hz page

Headphones for binaural. Keep the volume low and stop if anything feels uncomfortable; take care if you have tinnitus or sensitive hearing. Never while driving. Harmonance is not a medical device.

Twelve long gold ribbons of light on a deep violet ground, their waves lengthening gently from left to right
Twelve ribbons in alpha gold, drawn in code: each wave lengthens across the picture in the ratio of the session's beat, 12 falling to 9 a second.

What are alpha waves?

Alpha waves are a rhythm in the brain's electrical activity of 8 to 13 cycles a second, 8 to 13 Hz, recorded most strongly over the back of the head while you are awake. The glossary of the International Federation of Clinical Neurophysiology (IFCN), the clinical reference for EEG terms, says the alpha rhythm is "best seen with the eyes closed, during physical relaxation and relative mental inactivity", and "blocked or attenuated by attention, especially visual, and mental effort".

A single alpha wave lasts between 77 and 125 milliseconds, a thirteenth to an eighth of a second. EEG, electroencephalography, picks it up through electrodes on the scalp, which record the summed activity of millions of neurons beneath them.

Alpha is one of five bands named by speed. Delta, under 4 Hz, is strongest in deep sleep; theta runs from 4 to under 8 Hz; beta from above 13 to 30 Hz, as the glossary's correction puts it; and gamma above 30 Hz. The guide to brainwave frequencies takes each band in turn, and theta and delta, the two bands below alpha, have guides of their own.

Who discovered alpha waves?

Hans Berger made the first recording of a human EEG in 1924 and reported it in 1929, in a paper titled Über das Elektrenkephalogramm des Menschen, on the electroencephalogram of man. As Babiloni and colleagues recount, the first trace he published came from his son Klaus, aged 15, resting with his eyes closed, and Berger drew a 10 Hz sine wave beneath it to show that the main rhythm ran at about ten cycles a second. A year later he named it alpha.

Berger also saw what made it vanish. When his volunteers opened their eyes, took in something new or worked on a task, the alpha waves gave way to smaller, faster waves, which he called beta. The rhythm was widely accepted only after 1934, the turning point in Babiloni's account, when the British neurophysiologist Edgar Adrian and his colleague B. H. C. Matthews confirmed it in a paper they titled The Berger Rhythm. Adrian suggested it was generated locally in the visual cortex at the back of the brain, in relation to visual attention.

What happens to alpha waves when you close your eyes?

They grow, and by a lot. Harmonance measured how much in an open dataset, the EEG Motor Movement/Imagery Dataset on PhysioNet, recorded with the BCI2000 system. Before any tasks, each of its 109 volunteers sat for a one-minute baseline with eyes open and another with eyes closed. Harmonance took the three electrodes over the back of the head, O1, Oz and O2, worked out the power spectrum of each minute, and compared alpha power, the sum from 8 to 13 Hz, between the two. No recording was left out.

Of the 109 volunteers, 106 had more alpha with their eyes closed. For the median volunteer it was 3.5 times as much; a quarter had less than 2.2 times as much, and a quarter more than 8.5 times. At 10.25 Hz, where the median eyes-closed spectrum peaks, it stands 6.3 times higher than the eyes-open one: 141.5 against 22.5 µV² per hertz. The three exceptions had slightly less alpha with their eyes closed, between 0.90 and 0.95 times as much.

Median brain rhythm spectra of 109 volunteers: eyes closed, a gold curve peaking at 10.25 Hz; eyes open, a dashed curve that stays low
Closing the eyes multiplies alpha. In one-minute recordings from 109 volunteers, the median eyes-closed spectrum peaks at 10.25 Hz, 6.3 times higher than with the eyes open, and 106 of the 109 had more alpha with their eyes closed. Measured by Harmonance from the PhysioNet EEG Motor Movement/Imagery Dataset, electrodes O1, Oz and O2.
What the recordings show
MeasureResult
Volunteers109
More alpha with eyes closed106 of 109
Eyes-closed alpha over eyes-open, median3.5 times
The middle half of volunteers2.2 to 8.5 times
Median eyes-closed peak10.25 Hz
Power there, eyes closed141.5 µV² per Hz
Power there, eyes open22.5 µV² per Hz

Download the median spectra (CSV)

Download each volunteer's alpha (CSV)

Other studies find the same direction. In Barry and colleagues' study of 28 university students, alpha fell across the scalp when the eyes opened, while skin conductance, a common sign of arousal, rose. With the eyes closed, the students with more alpha had lower skin conductance. That fits the glossary's picture of alpha as a rhythm of relaxed wakefulness. It does not make alpha a measure of calm: visual attention blocks it, so opening the eyes lowers it whatever your mood.

Is everyone's alpha peak at 10 Hz?

No. Each person's alpha peaks at a frequency of their own. With their eyes closed, 107 of the 109 volunteers had a clear peak. The median sat at 10.25 Hz, a quarter of a hertz from the sine wave Berger drew, and the middle half of the peaks fell between 9.5 and 10.625 Hz. The lowest was at 7.75 Hz and the highest at 13 Hz.

Where 107 eyes-closed alpha peaks fell
Peak frequencyVolunteers
Below 9 Hz10
9 to 12 Hz94
Above 12 Hz3
The highest point between 7 and 14 Hz in each volunteer's eyes-closed minute, counted where it stood above both neighbours. Two of the 109 had no clear peak. Each volunteer's peak is in the CSV above.

A study by Haegens and colleagues recorded 51 volunteers with MEG, which reads the brain's magnetic fields, and found an average alpha peak of 10.3 Hz. The peak moved up as a memory task got harder, and the authors note that it rises into adulthood and falls again in older age. Their conclusion is that alpha works across a wider range than the 8 to 12 Hz band many studies use, so a fixed band can miss part of some people's alpha.

What do alpha waves do?

For a long time alpha was read as the brain idling. Jensen and Mazaheri set out why that reading falls short: several studies found alpha power rising with the number of items people held in memory, which is hardly rest. Their account, gating by inhibition, reads strong alpha as a sign that a region is being held back, so that information flows through the regions that are not. Alpha is also, in their words, "by far the strongest signal recorded by EEG and MEG".

Alpha turns up in research on ideas, too. A review by Fink and Benedek calls more alpha while people come up with ideas "among the most consistent findings in neuroscientific research on creativity", and reads it as attention turned inward. If you came here because ideas are not coming, Creative block? gathers tones people reach for at that moment.

None of this makes alpha a dial to turn up. These are patterns across groups of people, measured while they did particular tasks, and the studies here do not show that raising alpha with sound changes how calm or creative anyone feels.

Can binaural beats increase alpha waves?

A binaural beat is made by playing a slightly different steady tone to each ear. Nothing in the air beats: the brain hears the difference between the two tones as a slow pulse. H. W. Dove described it in 1839, as Chaieb and colleagues note, and Gerald Oster explained it for a wide readership in Scientific American in 1973: "Slow modulations called binaural beats are perceived when tones of different frequency are presented separately to each ear." Oster found the beat heard only when the tones sit below about 1,000 Hz, and the percept most likely forms in the brainstem, in the superior olivary nuclei, where the signals from the two ears first meet.

Two tones drawn at real speed: 108 and 120 Hz come back in step after 83 ms; after nine hours, 108 and 117 Hz take 111 ms
Where the session's beat comes from. At the start the right ear hears 120 waves a second to the left ear's 108, so the two tones drift apart and back in step 12 times a second; after nine hours the right ear is down to 117, and each drift takes 111 ms instead of 83. Drawn at the tones' real frequencies.
The session at its two ends
StartAfter nine hours
Left ear108 Hz108 Hz
Right ear120 Hz117 Hz
Beat12 Hz9 Hz
One beat lasts83 ms111 ms

Whether the brain's own rhythm follows the beat is another question. A systematic review by Ingendoh and colleagues found 14 EEG studies: five reported results in line with the idea, eight did not and one was mixed. For alpha, two studies found more alpha power and four found no difference.

Two small studies show what a null result looks like. Vernon and colleagues played a 10 Hz beat to one group and a 20 Hz beat to another, 22 people in all, for ten one-minute spells, and the EEG followed neither. López-Caballero and Escera played 14 people an 8.97 Hz beat on a 373 Hz tone, among beats in other bands, while they watched a silent film with subtitles, and found no change in the matching band.

One detail matters whenever a beat is said to have raised alpha. In most of the studies in the review, people listened with their eyes closed, and in the recordings above, closing the eyes alone multiplied alpha by 3.5 for the median volunteer. A fair test compares eyes-closed listening with the beat against eyes-closed rest without it.

Studies of how people do on tasks give a different picture. A meta-analysis by Garcia-Argibay and colleagues pooled 22 studies and 35 effect sizes and reported a medium average effect, g = 0.45, on tests of memory and attention and on other measures, varying with the beat, how long people listened and when. It pooled beats of many speeds, not alpha alone, and it does not show what a beat does to brain rhythms. Do binaural beats work? weighs those studies in full.

What is the Falling Alpha Session?

Harmonance's Falling Alpha Session is a nine-hour binaural beat made of two pure tones and nothing else. The left ear hears 108 Hz throughout. The right ear starts at 120 Hz and glides in one straight line to 117 Hz, so the beat starts at 12 Hz, in the upper half of the alpha band, and slows by 1 Hz every three hours: 11 Hz at three hours, 10 Hz at six and 9 Hz when it stops at nine. Over the whole session the two tones drift apart and back in step 340,200 times.

That span covers the range where most alpha peaks sit. In the recordings above, 94 of the 107 clear peaks fall between 9 and 12 Hz, and the beat passes the median peak, 10.25 Hz, five hours and 15 minutes in. That is arithmetic about the sound, not a mechanism: the studies above give no reason to expect the brain to follow it. Its 108 Hz tone sits well inside the range where a binaural beat can be heard, and the tones in the review's studies ran from 100 to 900 Hz.

The uploaded nine-hour master, measured
MomentRight ear, HzBeat, Hz
First second120.0012.00
Halfway, 4 h 30 min118.5010.50
Last second117.009.00
Read from the decoded audio of the uploaded master every five seconds, 6,481 readings in all. The left ear held 108.00 Hz throughout, and the largest difference from the planned line was 0.00038 Hz.

Harmonance measured the master itself, not only the recipe. Its loudness reads -23 LUFS, and its peaks stay more than 21 dB below the loudest a recording can go. For a beat that holds still instead, the 10 Hz alpha binaural beat plays a steady 10 Hz, and the Binaural Beats collection gathers all of the site's binaural beats and sessions.

How do you listen to an alpha session?

  • Use headphones. The beat exists only when each ear gets its own tone. Through a speaker the two tones mix in the air into an ordinary wobble in loudness, a monaural beat; isochronic tones and binaural beats explains the difference.
  • Keep it quiet. The US National Institute on Deafness and Other Communication Disorders warns that sounds "can be harmful when they are too loud, even for a brief time, or when they are both loud and long-lasting". Nine hours is long, so a level well below conversation is the sensible one.
  • Try it with your eyes closed, then open. Closing your eyes raises alpha by itself, as the chart shows. If you are curious whether the beat changes anything for you, sit for ten minutes with your eyes closed in silence, then ten with the session, and notice what differs.
  • Let it run in the background. Put it on for an unhurried afternoon or a slow evening, while you read, work or rest, and notice how it sits with you. If the beat feels intrusive, turn it down or stop. Harmonance Studio lets you layer tones into a session of your own, and Need to unwind? gathers gentle options in one place.

Sources and limits

The chart and the peak counts come from one open dataset: a minute with the eyes open and a minute with them closed from each of 109 volunteers, recorded before a set of movement tasks. A minute is short, and no recording or stretch of one was removed for blinks or muscle activity, so the figures are medians, which a few noisy recordings barely move. Each figure's numbers can be downloaded beside it. The EEG studies of alpha beats are few and small, and none tested this session. Berger's paper is in German, and the account of it here follows Babiloni and colleagues.

  1. Kane and colleagues, the IFCN glossary of terms used in clinical EEG: the alpha rhythm, its band and the other bands.
  2. Kane and colleagues, the glossary's correction: the beta band.
  3. Babiloni and colleagues, on Hans Berger and the alpha rhythm: the first recordings, Klaus's trace and the names alpha and beta.
  4. Berger, Über das Elektrenkephalogramm des Menschen: the first paper on the human EEG, in German.
  5. Adrian and Matthews, The Berger Rhythm, in Brain: the confirmation.
  6. Schalk, EEG Motor Movement/Imagery Dataset, version 1.0.0, PhysioNet: the recordings behind the chart, under the Open Data Commons Attribution licence.
  7. Schalk and colleagues, BCI2000: the system that recorded the dataset.
  8. Pollard and colleagues, PhysioNet as a global platform for biomedical research: PhysioNet's own citation.
  9. Barry and colleagues, EEG differences between eyes-closed and eyes-open resting conditions.
  10. Haegens and colleagues, variability in alpha peak frequency: 51 volunteers recorded with MEG.
  11. Jensen and Mazaheri, alpha activity and gating by inhibition.
  12. Fink and Benedek, EEG alpha power and creative ideation.
  13. Chaieb and colleagues, auditory beat stimulation: how a binaural beat is made and heard.
  14. Oster, Auditory Beats in the Brain, Scientific American: the classic description.
  15. Ingendoh and colleagues, binaural beats to entrain the brain?: the systematic review of EEG studies.
  16. López-Caballero and Escera, binaural beats and EEG power: an 8.97 Hz alpha beat among others.
  17. Vernon and colleagues, EEG changes and alpha and beta binaural beats.
  18. Garcia-Argibay and colleagues, a meta-analysis of binaural beats: memory, attention and other measures.
  19. NIDCD, noise-induced hearing loss: loud and long listening.

Tones to play while you read

Just be curious. This is not a medical device or medical advice. For any health concern, please speak to your doctor.