Gamma waves: attention, sound and the research

What are gamma waves?

Gamma waves are the fastest of the five classic rhythms in an electroencephalogram, or EEG, the recording of the brain's electrical activity from the scalp. The glossary that the International Federation of Clinical Neurophysiology publishes for EEG specialists sets the gamma band above 30 and up to 80 cycles a second.

The same glossary sets the slower bands: delta below 4 Hz, theta from 4 to under 8, alpha from 8 to 13 and, since its 2019 correction, beta above 13 and up to 30. Each band takes a Greek letter, and that letter is the only thing gamma waves share with the gamma rays of physics. The guide to brain wave frequencies walks through every band from delta to gamma.

What do gamma waves do?

Nobody can yet say for certain. A 2024 review sums up the mainstream view: gamma oscillations "aid in sensory information processing, attentional selection, and memory operations". Yet Buzsáki and Wang, in their review of how the rhythm arises, write that the functions and mechanisms of gamma rhythms "remain a matter of debate".

Their review is firmer on the machinery. Gamma comes from inhibition: fast inhibitory neurons act on the body of each neighbouring neuron, so that bursts of excitation and moments of restraint alternate in quick succession. The rhythm is short-lived, and its speed varies with the circuit that makes it.

Gamma also changes with practice. In a 2004 study, long-term Buddhist meditators produced sustained, high-amplitude gamma activity while they meditated, far more than a control group. The result is often quoted. It also runs straight into the problem of the next section.

Why are gamma waves hard to measure?

Muscles and eyes make electrical signals in the same range. In a 2007 experiment, two volunteers were recorded while a drug held their muscles completely still, and the power above 20 Hz on their scalp EEG fell by 10 to 200 times. The authors concluded that "most of the scalp EEG recording above 20 Hz is of EMG origin": muscle, not brain.

The eyes set a second trap. A 2008 study tracked the eyes alongside the EEG and found that the burst of gamma seen 200 to 300 milliseconds after a picture appears follows tiny involuntary eye movements, not a brain rhythm. The same authors add that true gamma oscillations have been shown conclusively by other methods. So a headline about gamma measured from the scalp, a consumer headband included, deserves one question first: were muscle and eye signals ruled out?

What happens when a sound repeats 40 times a second?

The brain keeps time with it, and this is the firmest result in the whole subject. In 1981, Galambos and colleagues reported that the brain's response to a brief sound looks like three or four cycles of a 40 Hz wave, and that "these waves combined to form a single, stable, composite wave" when sounds repeat at about 40 a second. Audiologists now call it the 40 Hz auditory steady-state response, and the paper already pointed to its use in hearing tests.

Keeping time is not the same as changing anything else. The response shows that the hearing system follows a 40 Hz rhythm. Whether following it does anything for attention or memory is a separate question, and the studies further down try to answer it.

Is there any 40 Hz in a 40 Hz binaural beat?

No. A binaural beat plays a different steady tone in each ear. The Harmonance 40 Hz gamma binaural beat plays 108 Hz in the left ear and 148 Hz in the right. The 40 is the difference between them, and it exists only where the signals from the two ears meet, inside the head.

Harmonance measured one minute of its own 40 Hz master recording, ear by ear. Each ear carries a single tone. At exactly 40 Hz the left ear's level is 275 dB below its own tone and the right ear's 186 dB below. That is digital silence: the smallest step a 24-bit file can take sits 117 dB below the tone. Each ear's loudness also holds steady, to within 0.00003 dB.

The numbers behind the chart

Measured, one minute of the recording	Left ear	Right ear
The one tone in the ear	108 Hz	148 Hz
Level at 40 Hz, below that tone	275 dB	186 dB
Smallest 24-bit step, below that tone	117 dB	117 dB
Rise and fall in loudness	0.00003 dB	0.00003 dB

What does change 40 times a second is timing. The two tones drift in and out of step, and the measured phase between the ears turns through a full cycle exactly 40 times a second. The brain compares the timing at the two ears to place sounds in space, and a binaural beat borrows that comparison. A 2005 study of the brain's 40 Hz response to binaural beats calls the beat "a signal that is not present in the stimulus, but generated within the brain".

Loudspeakers change the picture. Both ears then hear both tones, mixed in the air, and the combined sound swells and fades 40 times a second: an ordinary acoustic beat, also called a monaural beat. Even then the measured mix holds no tone at 40 Hz; the rhythm lives in the loudness. That is why a binaural beat needs headphones to stay binaural.

What does a 40 Hz binaural beat sound like?

Less like a beat than you might expect. At slow rates a binaural beat sounds like one tone wandering between the ears. A 2014 study that swept the rate from 3 to 60 Hz reports that "the percept changes to loudness fluctuation, roughness, and pitch with increasing beat rate". A 2023 systematic review puts the range in which binaural beats can be perceived at about 1 to 30 Hz. At 40 Hz, then, expect a quick flutter or a slight roughness in the sound, or simply two calm tones, rather than a slow pulse.

The brain still answers. The 2014 study recorded responses in the hearing cortex at every beat rate, and the largest at 40 Hz. The answer is small, though. The 2005 study found the binaural response weaker than the response to an acoustic beat at the same rate, and it vanished once the two tones rose above 3 kHz. A 2020 comparison found that binaural beats "weakly entrain the cortex when compared with monaural beat stimuli", and they moved none of its mood measures. Recordings from electrodes already placed inside the brain for clinical monitoring point the same way: the largest increases in power followed monaural 40 Hz beats. The guide to isochronic tones and binaural beats sets the kinds of beat side by side.

What did the 40 Hz studies in mice find?

Most of the excitement about 40 Hz comes from a series of mouse studies, and none of them used binaural beats. In 2016, Iaccarino and colleagues worked with mice bred to build up amyloid, the protein that forms plaques in the most common kind of dementia. A light flickering 40 times a second lowered two forms of amyloid-beta in the visual cortex of mice whose plaques had not yet formed.

In 2019 the same group turned to sound. Martorell and colleagues did not play a pair of tones. They played tones 1 ms long at 10 kHz, 40 a second at 60 dB, a train of clicks heard in both ears at once. Seven days of it improved the mice's spatial and recognition memory and lowered amyloid in the hearing cortex and the hippocampus. Only sound and light together lowered amyloid in the medial prefrontal cortex.

A 2024 study in the same kind of mice suggests how the clearing might work. The 40 Hz light and sound sped the flow of fluid through the brain, and "inhibiting glymphatic clearance abolished the removal of amyloid", meaning that when that flow was blocked, the amyloid stayed.

Other laboratories have not always seen the same thing. An independent team found in 2023 that 40 Hz flickering light "did not engage native gamma oscillations" in the brain regions they recorded, that the mice avoided the flicker, and that plaques did not change reliably. The 2024 review notes that several studies have failed to replicate the original results. A mouse study cannot tell us what a sound does for a person listening at home, and these studies played clicks and light, not two steady tones.

What have the 40 Hz trials in people shown?

Early safety, and no clear effect yet on the main measures. The best-known trials in people all used a device that delivers 40 Hz light and sound together for an hour a day. None used binaural beats.

A three-month pilot in 15 people with mild dementia found the stimulation well tolerated, with exploratory signs in its favour: less shrinking of the hippocampus, less widening of the brain's fluid spaces and better recall of faces paired with names. The larger OVERTURE trial gave 76 people with mild to moderate dementia an active or a sham device for six months. It found no unexpected serious side effects and "no statistical separation between active and sham" on its main memory and thinking scales. A few secondary measures, among them daily living and whole brain volume, favoured the active group at nominal significance, a weaker standard that does not correct for testing many measures, and the trial was not designed to show whether the device works. Five people from the pilot carried on for two years of daily use with no adverse events reported, far too few for any conclusion.

The largest trial, HOPE, enrolled 673 people for about a year of daily use. The trial registry lists it as completed, with no results posted yet. OVERTURE and HOPE were both run by the company that makes the device.

Do 40 Hz binaural beats change attention or memory?

The studies point both ways, and most are small. On one side, 25 students made fewer errors on an attention task while hearing a 40 Hz binaural beat than while hearing pink noise, though their recordings "could not confirm the hypothesized neural entrainment". In a three-day training study, 40 Hz beats drew a strong gamma response and both groups improved by the same amount overall; the authors read the order of the gains as faster learning with 40 Hz. After 20 minutes of a 40 Hz binaural beat, volunteers in a 2017 study recalled more words from the working memory part of a list.

On the other side, 58 adults showed no significant differences in reaction time, errors or attention between a 40 Hz binaural beat and a plain control tone. Thirty people showed no significant changes in memory scores or in their EEG. In 28 people, a gamma binaural beat and plain white noise both raised gamma power, a reminder that any sound can move the EEG and that a fair study needs a comparison sound. And a 2017 study of gamma binaural beats and how the mind binds the features of what it sees was retracted by its journal.

Pooled results do not settle it. A 2019 meta-analysis of 22 studies found a medium overall effect for binaural beats, but it mixed many beat rates and outcomes, so it says little about 40 Hz alone. The 2023 review of EEG studies counted five results in line with the idea that the brain locks to the beat, eight against it and one mixed. Taken together: a few small positive results, several careful null results, and no large trial of 40 Hz binaural beats. The guide to what binaural beat studies found covers the other beat rates.

Is it safe to listen to gamma waves?

You cannot hear a gamma wave itself; what plays is a sound with a 40 Hz rhythm in it. A 40 Hz binaural beat is two quiet, steady tones, so the main thing to watch is volume. The World Health Organization's advice for headphones is to keep the volume down, no more than 60 per cent of maximum on your device. The trials of 40 Hz light and sound reported no unexpected serious side effects at an hour a day, though those devices also flash light, which a binaural beat does not.

How can you listen to a 40 Hz binaural beat?

With headphones, at a low volume. The 40 Hz gamma binaural beat plays 108 Hz in the left ear and 148 Hz in the right, free and with no account. The rising gamma session starts at a 32 Hz beat and climbs slowly to 40. The binaural beats collection holds a beat for every band from delta to gamma, and if you arrive with a feeling rather than a number, How do you feel? starts there.

Press play, let the sound settle for a few minutes and notice what you notice. The other bands have their own guides: alpha waves, theta waves and delta waves.

Sources and limits

The measurements of the 40 Hz recording are Harmonance's own, and the data sits beside the chart. The studies below concern 40 Hz sound, light and binaural beats in general; none tested this recording. Results in mice do not carry over to people on their own, and the largest trial in people has not yet reported.

1. Kane and colleagues (2017), the revised glossary of EEG terms: the gamma band and the other bands.
2. Deng and colleagues (2024), a review of gamma oscillations and gamma stimulation: what gamma is linked with, and the failed replications.
3. Buzsáki and Wang (2012), mechanisms of gamma oscillations: how the rhythm arises, and the open debate.
4. Lutz and colleagues (2004), gamma in long-term meditators.
5. Whitham and colleagues (2007), scalp recording during paralysis: muscle above 20 Hz.
6. Yuval-Greenberg and colleagues (2008), gamma bursts and tiny eye movements.
7. Galambos and colleagues (1981), a 40 Hz response to sound recorded from the scalp.
8. Ross and colleagues (2014), responses to slow and fast binaural beats: what the beat sounds like at each rate.
9. Ingendoh and colleagues (2023), a systematic review of binaural beats and the EEG.
10. Schwarz and Taylor (2005), steady-state responses to binaural and monaural beats.
11. Orozco Perez and colleagues (2020), binaural beats through the hearing pathway.
12. Becher and colleagues (2015), recordings inside the brain during monaural and binaural beats.
13. Iaccarino and colleagues (2016), 40 Hz flicker and amyloid in mice.
14. Martorell and colleagues (2019), 40 Hz sound and light in mice: the click train and the light pulses.
15. Murdock and colleagues (2024), 40 Hz stimulation and the brain's fluid flow in mice.
16. Soula and colleagues (2023), 40 Hz light without native gamma in mice.
17. Chan and colleagues (2022), feasibility and pilot studies of 40 Hz light and sound.
18. Hajós and colleagues (2024), the OVERTURE trial.
19. Chan and colleagues (2025), two years of daily use in five people.
20. ClinicalTrials.gov, the HOPE study record.
21. Engelbregt and colleagues (2021), binaural and monaural beats, attention and the EEG.
22. Ross and Lopez (2020), 40 Hz binaural beats during attention training.
23. Jirakittayakorn and Wongsawat (2017), a 40 Hz binaural beat, memory and mood.
24. Leistiko and colleagues (2023), gamma binaural beats and attention.
25. Borges and colleagues (2023), gamma binaural beats, memory and the EEG.
26. Dos Anjos and colleagues (2024), EEG changes during beats and white noise.
27. Experimental Brain Research (2021), the retraction note for a 2017 gamma binaural beat study.
28. Garcia-Argibay and colleagues (2019), a meta-analysis of binaural beats.
29. World Health Organization, safe listening: keeping the volume down.