Billions of neurons fire in sync, and electrodes on the scalp pick this up as waves. The frequency of those waves tells you whether you are asleep, daydreaming or solving a problem. Tap any channel of the recording to explore a rhythm.
2 s window · schematic signalsthe higher the frequency, the smaller the amplitude
0.5 Hz481330100 Hz
Lab
Tune the frequency
Move the slider and see which rhythm the frequency falls into. With headphones you can hear a binaural beat: 200 Hz in the left ear, slightly higher in the right, and the brain hears a “beat” at the difference.
10 HzAlpharelaxed wakefulness
This is a demonstration, not a treatment. Research on binaural beats is mixed: any effect on attention or anxiety is small. Start at a low volume.
Mixer · compared with a real night
Build your own EEG
Real EEG is a mix of all rhythms at once. Move the sliders and watch the signal change. The site compares your mix with how the rhythms were distributed in each stage of a real night recording.
As in the recording:
Closest match—
Position 50 means the rhythm’s average level across the whole night. The site looks at which rhythms you made stronger or weaker than average and compares that with each sleep stage. Alpha comes from the occipital electrode, the other rhythms from the frontal one. There is no gamma in the recording: the device recorded up to 50 Hz, so gamma does not affect the comparison.
24 hours
Which rhythm leads through the day
All five rhythms are always present at once. Only which one is loudest changes. Here is a typical day of someone who sleeps from 23:00 to 7:00. You can see a real night recording in the “Science” section.
00040812162024
No mysticism
What is worth knowing
You can’t “switch on” a rhythm with a button
You can shift the balance: closing your eyes boosts alpha within seconds, a hard task raises beta. It is a consequence of your state, not a lever.
There is no good or bad rhythm
Beta is needed for work, but too much high beta is linked to anxiety. Daytime theta in an adult can mean fatigue. What matters is the right balance for the task.
40 Hz gamma and Alzheimer’s
In mice, light and sound flickering at 40 Hz reduced amyloid plaques. Clinical trials in humans are under way, with no conclusions yet.
The tools that work are boring
Sleep, movement, breathing, meditation and an engaging task change your rhythms more reliably than any tracks or gadgets.
Where rhythms are born
Inside the brain
Each rhythm has its own “places”. The model shows where scientists found them and how they behave. Rotate the brain with your finger or mouse and switch between rhythms.
The international 10–20 system (1958): electrodes are placed at 10% and 20% of the distances between the bridge of the nose, the bump at the back of the skull and the ears. Highlighted in red are the pairs used for the night recording in the “Science” section: Fpz-Cz (forehead to crown) and Pz-Oz (top to back of the head). The device records the voltage difference between two points.
Delta · 0.5–4 Hzslowed down
↻ drag to rotate
Scale
What EEG actually sees
One electrode hears millions of neurons
A signal appears on the scalp only when tens of thousands of cortical neurons work in sync, over roughly a square centimetre or more. EEG cannot see a single cell.
Microvolts through bone
A typical EEG amplitude is 10 to 100 µV. That is a million times less than the voltage of a battery. The skull and skin blur the signal.
Depth has to be inferred
Hippocampal theta is recorded with electrodes inside the brain: in rats, and in people with epilepsy before surgery. It is almost invisible from the surface.
Data · papers · models
The science of rhythms
Seven stories about how scientists learned what the other sections describe. Every chart is labelled with where its data come from.
Real recordingNumbers from the paperModel of the effect
150 years in a minute
How brain rhythms were discovered
Sleep · 1
One night of one person
Real recording
This is real EEG: a 7-hour night recording of a 33-year-old woman (about 5.5 hours of it asleep), electrode on the forehead (Fpz-Cz), 100 samples per second. At the top are the sleep stages a clinician scored in 30-second epochs. Below is a spectrogram: the brighter, the stronger the frequency. You can see that sleep runs in cycles. In the first half of the night delta glows brightly at the bottom; towards morning there is less and less deep sleep.
Hover or swipe across the chart
Listen to this night
The signal is played hundreds of times faster, so brain waves become sound. Deep sleep is a low hum, light sleep with spindles whistles, wakefulness rustles. Best with headphones; start at a low volume.
Clock face: the night runs around the dial like a clock. The inner ring shows stages (closer to the centre means deeper sleep), the outer ring shows rhythm shares. Numbers mark sleep cycles, each ending with an episode of REM sleep.
8 seconds of each stage
Data: Sleep-EDF Expanded, recording SC4001, PhysioNet; Kemp et al., 2000, IEEE Trans Biomed Eng. Sleep cycles described by Dement & Kleitman, 1957, EEG Clin Neurophysiol. Stages 3 and 4 are merged into N3 under current rules.
Sleep · 2
How alpha fades as we fall asleep
Real recording
The same night, but from an electrode over the back of the head (Pz-Oz). The person lies with eyes closed, not yet asleep: alpha is strong. At sleep onset it collapses and theta takes its place. This is the rhythm Hans Berger saw first and named alpha.
Share of alpha and theta, first 75 minutes
alpha 8–13 Hztheta 4–8 Hz
Alpha before sleep, 8 seconds
You can see “spindles” of waves at about 10 per second. If the person opened their eyes, they would almost vanish: this is the Berger effect.
Data: Sleep-EDF SC4001, channel Pz-Oz. Discovery of the alpha rhythm: Berger, 1929, Archiv für Psychiatrie.
Sleep · 3
Does the brain wash itself during sleep?
Numbers from the paper
In 2013 Maiken Nedergaard’s group measured the space between brain cells in mice. During sleep it expanded by about 60%, fluid flowed more freely and carried away metabolic waste. Beta-amyloid, a protein linked to Alzheimer’s disease, was cleared about twice as fast during sleep.
Diagram: cells shrink, the gaps between them grow.
Interstitial space, share of volume
Awake
~14%
Asleep
~23%
+60%interstitial space volume during sleep
×2clearance rate of beta-amyloid
These are mice. In humans a similar mechanism is studied with MRI, and the exact numbers differ.
Scientific debate
In 2024 the group of Nicholas Franks and William Wisden measured dye clearance from mouse brains directly. In sleeping mice it was cleared about 30% more slowly than in awake mice, and about 50% more slowly under anaesthesia. This directly contradicts the 2013 conclusion. Supporters of the hypothesis say the difference comes from the method. The question is open for now, but nobody disputes that sleep matters for memory and attention.
Xie et al., 2013, «Sleep drives metabolite clearance from the adult brain», Science 342:373. Counter-evidence: Miao et al., 2024, “Brain clearance is reduced during sleep and anesthesia”, Nat Neurosci 27:1046.
Sleep · 4
A click in time with the slow wave boosts memory
Numbers from the paper
In the evening, 11 volunteers learned 120 word pairs. At night a computer watched their EEG and, after catching the trough of a slow wave, played a soft click of pink noise half a second later, right on its crest. The waves grew larger, and in the morning people recalled more words than after a night without clicks.
When the click plays
Wave crest: cortical neurons are active. The click lands on the rise. The same sound played out of phase had no effect.
Overnight gain in recalled word pairs
No clicks
13.0
With clicks
22.2
11participants, each in both conditions
+9%slow-wave amplitude
50 msclick length, 55 dB
Ngo et al., 2013, «Auditory closed-loop stimulation of the sleep slow oscillation enhances memory», Neuron 78:545.
Meditation · 5
Gamma in monks with 10,000 hours of practice
Numbers from the paper
Richard Davidson and Antoine Lutz recorded EEG from Tibetan monks during unconditional compassion meditation and compared them with students who had been taught the same practice for a week. The monks had higher gamma even at rest, and during meditation it rose to levels never before recorded in healthy people.
8long-term practitioners
10–50 thousandhours of meditation over 15–40 years
10novices, one week of training
25–42 Hzgamma band in the paper
Ratio of gamma to slow rhythms
Rest
Meditation
Novices
baseline
almost unchanged
Monks
higher than novices
sharp rise
Levels are shown qualitatively, without exact values. A limitation of the study: the monks were older than the students, and the groups differed in more than experience.
Lutz et al., 2004, «Long-term meditators self-induce high-amplitude gamma synchrony during mental practice», PNAS 101:16369. Review of meditation and EEG: Cahn & Polich, 2006, Psychol Bull.
Memory · 6
Place cells fire to the rhythm of theta
Model of the effect
John O’Keefe found neurons in the rat hippocampus that switch on only in a particular spot of a room, and won the 2014 Nobel Prize for it. Later he noticed that as the rat runs through “its” spot, the neuron fires at an ever earlier phase of each theta wave. Time is encoded by rhythm.
theta ≈ 8 Hz in ratsplace-cell spike
The animation is slowed 4× and based on a description of the effect, not on a recording.
O’Keefe & Recce, 1993, «Phase relationship between hippocampal place units and the EEG theta rhythm», Hippocampus 3:317.
Memory · 7
The more you hold in mind, the stronger frontal theta
Model of the effect
Participants were shown 1, 3, 5 or 7 digits and asked to hold them for 3 seconds. A magnetoencephalograph picked up a 7–8.5 Hz rhythm over the frontal cortex (on the border of theta and alpha), and its power grew with every added digit. Try it yourself: choose the load.
The amplitude in the diagram is illustrative. The paper had 10 participants, and the rise in theta power with load was systematic.
Jensen & Tesche, 2002, «Frontal theta activity in humans increases with memory load in a working memory task», Eur J Neurosci 15:1395.
Breathing and meditation timer
Practice
Choose which rhythm you want to support and for how long. The circle shows when to breathe in and out. Each technique is based on research, with links below the timer.
Readypress start
Time
If you feel dizzy, return to normal breathing. Breath holds are not recommended during pregnancy or with heart or lung conditions without a doctor’s advice.
From day one to 90 years
Rhythms and age
A baby’s brain runs slowly, by age ten the main rhythm speeds up to adult alpha, and with age it slows down again. Sleep changes too: there is less deep sleep every decade. Move the age slider and see what happens.
30 yearsadult
Main rhythm over the back of the head
≈10 Hzadult alpha
2-second window, frequency at real scale.
Sleep at this age
Maturation
How the main rhythm speeds up
Clinical norms
Dots are clinical EEG norms for children: 3–4 Hz at 2–3 months, 6 Hz at one year, 8 Hz at 3 years, 9 Hz at 8 years, 10 Hz by age 10. After young adulthood alpha frequency slowly declines, and the curve looks like an inverted U. How much it drops varies between studies, so that part is dashed.
Besides rhythms, EEG has a background: the higher the frequency, the weaker the signal. On a log plot this is almost a straight line. Bradley Voytek showed that in older people it is flatter. This means neurons fire less coherently, and there is more “noise” in the signal.
11 and 13younger and older participants
r 0.57 → 0.36correlation of age and working memory once noise is accounted for
Noise statistically explained a substantial part of the age-related decline in working memory. The spectral slope in the drawing is schematic.
Voytek et al., 2015, «Age-related changes in 1/f neural electrophysiological noise», J Neurosci 35:13257
Dementia prevention · research
How to slow brain ageing
Ageing can’t be cancelled, but its pace can be influenced. Here is only what has been tested on thousands of people. The label on each block shows the strength of evidence: a controlled trial or an observational study.
Trial (RCT)ObservationalReview
The main risk map
45% of dementia cases are linked to 14 factors we can change
Review
The Lancet Commission estimated what share of dementia cases worldwide is linked to each factor. If all 14 disappeared, almost half of cases could in theory be prevented.
These are population-level shares, not personal risk. The factors overlap, so you can’t add them up for one person.
Livingston et al., 2024, «Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission», Lancet 404:572
What has been tested
Five trials worth knowing
Walking grows the hippocampus
RCT
Hippocampal volume over one year
Walking
+2.0%
Stretching
−1.4%
Bar length shows the size of the change, not its sign.
120 older adults walked briskly 3 times a week for 40 minutes for a year. Their hippocampus, the memory centre, grew by 2%. That roughly reversed 1–2 years of age-related loss. In the stretching group the hippocampus kept shrinking. Spatial memory improved in both groups, but among walkers the volume gain was linked to better memory.
Everything at once works better: the Finnish FINGER trial
RCT
How much more thinking improved than in the control group
Overall score
+25%
Executive function
+83%
Processing speed
+150%
Memory
n.s.
1,190 people aged 60–77 at elevated dementia risk spent two years changing their diet, exercising (cardio and strength), doing cognitive training and managing blood pressure, blood sugar and cholesterol. Compared with a group that received only advice, thinking improved more. The effect on memory was not significant (“n.s.”).
Blood pressure below 120: −19% risk of cognitive impairment
In the SPRINT MIND trial, people over 50 were treated to a systolic pressure of about 120 instead of 140. Over 5 years, mild cognitive impairment developed 19% less often. The difference for dementia did not reach significance.
Hearing aids: −48% rate of decline, but not for everyone
977 people aged 70–84 with hearing loss, for three years. Overall there was no effect. Among those with more risk factors, thinking declined 48% more slowly.
Sleeping 6 hours or less at 50–60: about 30% higher risk
Almost 8,000 Britons were followed for 25 years. Those who consistently slept little in midlife developed dementia more often. This is an association, not proven cause.
604 people over 65, for three years. Thinking improved both on the MIND diet and in the control group with the same mild calorie reduction. No difference between them was found. So what matters is more likely healthy eating overall than a special list of foods.
This is a review of research, not medical advice. Have your blood pressure, cholesterol, hearing and vision checked by a doctor.
Memory · attention · clarity
How to improve brain performance
Not “hacking” with supplements, but what in experiments noticeably changed results today or within a few weeks.
Sleep · 1
6 hours of sleep for two weeks is like a night without sleep, and you don’t notice
RCT
48adults aged 21–38 in the lab
14 daysof 4, 6 or 8 hours in bed
1 and 2 nightswithout sleep — the same number of attention lapses as the 6- and 4-hour groups
Attention and working memory in the 6-hour group fell day after day. Yet subjective sleepiness barely rose after the first days. People got used to the feeling, but not to its consequences. Only the 8-hour group held steady.
Students learned a text in three ways: read it 4 times (SSSS), read it 3 times and recall once (SSST), read once and recall from memory three times (STTT). Switch when to test.
The irony of the experiment: the rereaders were the most confident they would remember (4.8 out of 7 vs 4.0), and recalled the least a week later.
Roediger & Karpicke, 2006, «Test-enhanced learning», Psychol Sci 17:249. Spacing between repetitions: Cepeda et al., 2006, meta-analysis of 317 experiments, Psychol Bull 132:354.
Attention · 3
A walk in the park restores attention
Experiment
Gain on the backwards digit span test
Park
+1.5
City
+0.5
Digits a person can hold and repeat in reverse order. Before the walk, 7.9 on average.
38 students walked for 50–55 minutes: once through an arboretum, a week later through downtown, or the other way round. After the park, attention improved noticeably; after the city, hardly at all. Mood also improved after the park, but the attention gain was not linked to it.
NASA pilots on long-haul flights were given 40 minutes to rest and slept 26 on average. In the last hours of the flight their alertness and reactions were better than those of pilots without rest. For a daytime nap, 10–20 minutes is usually recommended: after longer sleep you can wake up groggy.
Improves alertness and reaction speed, especially when you are short on sleep. It boosts complex thinking less. Late coffee disrupts sleep and eats up the gain the next day.
People who constantly keep lots of media open were worse at filtering out distractions and slower to switch. Replications give mixed results, but turning notifications off certainly won’t hurt.