A plain-language experimental protocol for testing whether auditory entrainment stimulus produces a measurable change in the EEG. Exact settings, a control condition, and — the part that usually goes missing — an explicit statement of what a null result looks like.
Brainwave entrainment has a large popular literature and a small experimental one. Claims circulate faster than replications, and a great many of the frequency values in circulation have no traceable origin at all.
This protocol does not argue for a conclusion. It is a recipe, written so that somebody with access to an EEG can run it without having to reconstruct the design first, and so that two people running it independently are running the same thing.
Who this is for: anyone with access to an EEG — a university lab, a psychology or neuroscience department, a physiology teaching lab, a maker space, a student with a supervisor. Undergraduate coursework is a perfectly good venue for this.
The author of this protocol does not own an EEG and has not run it. That is precisely why it is published.
Hypothesis (H1). Listening to an amplitude- or binaurally-modulated tone at a fixed beat frequency f produces a measurable increase in EEG spectral power at f (and/or its harmonics), relative to a control condition using the same carrier tones with no beat, in the same participant on the same day.
That is deliberately narrow. It is about a measurement, not about how anyone feels, sleeps, focuses or performs. Nothing on this page tests, claims, or implies any health, wellbeing, cognitive or therapeutic outcome, and no such outcome should be inferred from any result obtained with it.
Set these explicitly rather than loading a preset, so that what you report is what you ran. Read every value back off the screen before you start.
| Condition | Mode | Carrier L | Carrier R | Beat | Wave |
|---|---|---|---|---|---|
| A — binaural | Binaural | 196.085 Hz | 203.915 Hz | 7.83 Hz | Sine |
| B — isochronic | Isochronic | 200.00 Hz | 200.00 Hz | 7.83 Hz | Sine |
| C — control | Binaural | 200.00 Hz | 200.00 Hz | 0.00 Hz | Sine |
In a binaural condition the generator splits the beat across the two carriers — set carrier 200.00 Hz and beat 7.83 Hz and the app will show you 196.085 Hz left and 203.915 Hz right. Verify those two numbers on screen; they are the actual stimulus.
Condition C is the control and it is the whole experiment. It is acoustically matched — same carrier, same level, same duration, same instruction to the participant — and differs only in carrying no beat. Without it you cannot separate an entrainment effect from the effect of sitting still with your eyes closed listening to a tone.
Fixed across all conditions: level constant (no fade in or out), continuous mode, no session timer, one voice only, volume set once and never touched again.
7.83 Hz is used here because it is a commonly cited value and sits inside the alpha/theta boundary region, not because it is special. The protocol works with any beat frequency; if you change it, report the value you used.
Repeat the whole session on a second day if you can. Within-participant reliability across days is more informative than adding participants.
State this before you collect anything. It is the part most write-ups omit, and it is the reason this page exists.
The result is null if:
A null result here is a real finding and it is worth publishing. The absence of published nulls is a large part of why this field is in the state it is in.
Run it under whatever ethical approval your institution requires. Auditory stimulus of this kind is ordinary listening at ordinary volume, but participant consent and institutional review are yours to obtain, not ours to waive.
If you use preset values from any generator, ask where they came from before you report them. In FreqCraft, seven of the single-voice presets come from a compiled frequency listing by Michael Triggs, and two of the five-voice stacks come from an archived product listing recovered in 2019 carrying carrier and beat values only. One preset, Awakened Mind, is named after C. Maxwell Cade and Nona Coxhead's 1979 book and states inside the app which of its values were measured by spectrum analysis and which were interpreted.
That last distinction is the one worth carrying into your own write-up: separate what was measured from what was inferred, and say which is which.
Including nothing. Preregister if you can — OSF is free. Post the raw recordings and your analysis script alongside the result if your institution allows it, so the next person can check the pipeline rather than take your word for it.
This protocol is released into the public domain. No attribution required, no permission needed, no strings. Change it, improve it, disagree with it in print.