The research

The binaural beat frequency chart, and which rows are real

12 August 2026

Every binaural beats site has this table. Here it is, with the labels you will recognise, and then the part those sites leave out: which rows are findings and which are folklore.

The chart

The band column is real. It is a naming convention from EEG research, and the boundaries are the ones this app uses: delta 0.5–4 Hz, theta 4–8, alpha 8–13, beta 13–30, gamma 30–100. A 4 Hz beat sits on the delta/theta boundary and is conventionally filed as theta.

That reattachment is the entire commercial premise of this field, and it is the part with the least behind it.

Why the chart cannot mean what it appears to

For "play 2 Hz, get deep sleep" to work, a beat at a given rate would have to reliably drive cortical activity to that rate. That is the entrainment hypothesis.

Ingendoh 2023 reviewed fourteen studies of EEG entrainment. Five supported it. Eight contradicted it. One was mixed.

Two more, testing rates directly on the chart above. Goodin 2012 tested 7 Hz theta and 16 Hz beta against a control and found no frequency-specific increase in EEG power. Lopez-Caballero and Escera 2017 tested beats across several rates and failed to enhance either EEG power or emotional arousal, and reported the null result as a null result.

So the mapping the chart implies is not supported, and two studies tested rows of it specifically and found nothing.

The three rows that are real

Three rates in this app have an experiment behind that exact rate. Not the band — the rate.

3 Hz. Jirakittayakorn and Wongsawat 2018 played a 3 Hz beat during sleep and measured more delta activity and longer stage N3. A small experiment, and not an insomnia trial.

6 Hz. Jirakittayakorn and Wongsawat 2017 found a 6 Hz beat on a 250 Hz carrier produced more theta, including the frontal-midline theta associated with focused attention. An EEG resemblance, which is not the state itself.

15 Hz. Beauchene 2017 found a 15 Hz beat improved verbal working memory accuracy and changed cortical connectivity. Small, and cognitive findings elsewhere are inconsistent.

Note that 15 Hz is not even on the standard chart, which jumps 12 to 16. The tested rate and the marketed rate are not the same number.

Gamma has one recent result worth naming: Melnichuk, Cooper and Hawk 2025 tested gamma beats on 80 students and found improved attention with EEG-confirmed entrainment, varying by carrier, with no effect on vigilance decline. Exploratory, and the most interesting open thread in the field.

What that leaves

Of ten rows, three have direct support, two have been tested and found nothing, and the rest have never been tested as rates at all. The mood words on all ten are borrowed.

That is a much smaller claim than the chart makes, and we would rather print the smaller true one.

How to actually pick a rate

Since the chart will not do it for you:

Pick by what you are doing, not by a number. The bands are a rough map of a contested theory. Slow rates for settling, faster for working, is about as much as the evidence carries.

If you want the tested ones, they are 3, 6 and 15 Hz. Not because those are magic, but because someone actually ran the experiment.

Give it twenty minutes. The sleep and attention experiments ran sessions of that order. Nothing suggests anything happens in two.

Use headphones, or there is no binaural beat at all - on a speaker the channels mix in the air before reaching either ear.

And notice whether it works for you. Given how much the trials disagree with each other, your own experience over a few sessions is genuinely competitive evidence with the literature for the question of whether you should use this.

Every study named here is on the research page with its sample size, effect size and what undercuts it.

All posts