Monaural Beats Two tones, one pulse

Two tones, close in pitch, added together before they reach your ear. That is the entire recipe.

Play 200 Hz and 210 Hz through the same speaker and the two waveforms slide in and out of step. Where their peaks coincide, the sound is loud. Fifty milliseconds later they are opposed and pulling against each other. A hundred milliseconds after the first loud moment they line up again, which gives ten loud moments a second. The beat rate is 10 Hz because the tones are 10 Hz apart. Nothing more elaborate is going on.

If both tones sit at the same level, the cancellation halfway through each cycle is complete and the sound drops to nothing before swelling back. So the pulse is not a perceptual trick. It is in the electrical signal, in the file, and in the air. Put a microphone in front of the speaker and the rise and fall shows up on the recorded waveform.

Run that file through a spectrum analyser and you find exactly two things: a line at 200 Hz and a line at 210 Hz. There is no 10 Hz component. Nothing sits at 205 Hz either, though 205 is roughly the pitch you hear. The pulse is what the two tones do to each other over time, not a frequency present in the sound. That is why a phone speaker with no output at all below 150 Hz can carry a 2 Hz monaural beat without difficulty. It never has to move at 2 Hz; it only has to play two tones.

None of this was invented for brainwave audio. Beating is one of the oldest measuring tools in music.

Two piano strings meant to sound the same note, one of them slightly sharp, produce an audible wobble. A tuner does not have to judge pitch in the abstract. They count the wobble, tighten the string, and listen to the wobble slow down. When it stops entirely the two strings are at the same frequency, because a beat rate of zero means a frequency difference of zero. Guitarists do a rough version of this by ear. A physics demonstration does it with two tuning forks and a lump of wax on one prong.

Organ builders went the other way and used it on purpose. A celeste rank is deliberately tuned slightly off its neighbour so the two pipes beat against each other, which is where that shimmering, breathing quality comes from.

A monaural beat is the same phenomenon with the numbers chosen so the wobble lands inside a brainwave band. It also inherits the tuner's problem in reverse. Down at the bottom of delta the two tones are a hertz or so apart, one swell every second, and a beat that slow is hard to tell from a tone that is not beating at all.

The three formats sold as entrainment audio differ in one respect that decides everything else about them: where the rhythm physically is.

You can build a binaural beat out of the same two tones and hear no beat at all. Send 200 Hz to one ear and 210 Hz to the other and neither channel is modulated; measure either one on its own and it is a plain sine wave. The beat arrives only after the two ears' signals have been combined, which Schwarz and Taylor in 2005 described as a signal "that is not present in the stimulus, but generated within the brain". Sum those same two tones in the air and the beat is simply there. Headphones are not a recommendation for the binaural version. Without them there is nothing to hear.

An isochronic tone is a single tone switched on and off, so the rhythm is in the audio as literal gaps.

A monaural beat sits between the two. The modulation is real and measurable like the isochronic case, but continuous: the level glides up and down instead of being gated, so there are no gaps, no switching edges, and no click to remove.

Play a binaural file through speakers and both tones reach both ears through the air, so they sum acoustically before you hear them. What arrives at your ears is an ordinary acoustic beat at that rate, plus whatever the room adds. Not a degraded binaural beat. A different stimulus.

200 and 202 Hz: one note breathing, a swell every half-second you can watch arrive. 200 and 240 Hz: a rough edge on a note. Same format, same carrier, and there is no way to get the first sound at the second rate.

That constraint belongs to this format alone. An isochronic pulse is a tone multiplied by a rhythm, so pitch and rate are separate dials: gate 400 Hz at 2 Hz or at 20 Hz and it is the same 400 Hz tone either way. A binaural beat disturbs the audio even less, since each channel stays one steady tone no matter what rate you ask for. In a monaural beat the rate is the spacing between the two tones. Every change to the rate is a change to the sound, and there is no dial to hold one still while you move the other.

Walk it up from the bottom. A hertz or two apart and the pair reads as a single note with a slow tremor in it; the slower you go, the more it is simply a tone. Around 10 Hz the flutter is unmistakable and countable. Higher, the swells stop resolving as separate events and fuse into a texture, somewhere in the mid-to-high teens, for the reasons the isochronic page sets out. A 25 Hz beta beat is not twenty-five perceptible anythings. It is 200 against 225 Hz, and it sounds like one restless, slightly sour note. Widen the gap further and the two tones begin leaving each other's patch of the cochlea, the roughness thins out, and what is left is a musical interval: two notes, no beat.

None of that is a fault in the file. It is the ear.

So the format has a comfortable range and an uncomfortable one. Below the teens you can hear what you selected. Above the teens you are choosing an acoustic modulation rate on trust, because the perceptual evidence of it has gone. Arkhitec's monaural rates run from 1 Hz to 50 Hz, covering all five bands, and the top of that range is honest about what it is: a 40 Hz preset is 200 Hz against 240 Hz, a microphone would record the modulation, and nobody hears forty events a second. Take it if an acoustic 40 Hz envelope with no headphones is what you want.

The acoustics above are settled. What the beat does to a brain is not.

The clearest laboratory look at a monaural beat sits inside a paper that was not about monaural beats. Schwarz and Taylor, Clinical Neurophysiology 2005, were chasing the binaural case: a tone to each ear, 40 Hz apart, and a hunt through the EEG for activity following that 40 Hz difference. They needed something to hold the result against, so they also fed both tones into one ear. That is a monaural beat, and in their design it was the control.

The binaural response turned up at a low mean stimulus frequency near 400 Hz and had gone by 3 kHz. The control response was larger, and it was there at low and high frequencies alike.

That gets stretched a long way past what it says. An auditory steady-state response is scalp-recorded electrical activity following the envelope of a sound. It tells you the auditory system tracked the modulation. It is not sleep quality, attention or mood, and nobody has established what a larger response is worth to the person having it.

Control condition is roughly the monaural beat's standing in the literature, and the reason has nothing to do with results. A binaural beat is a puzzle. Hearing something present in neither ear's signal is direct evidence that the two ears are combined centrally, and a puzzle is worth an experiment. Two tones interfering was described and understood long before anyone put an electrode on a scalp. There was no question to answer. A thin literature here records disinterest, not defeat, and it is not evidence in favour either.

Entrainment research generally has not settled it either. Ingendoh, Posny and Heine's 2023 PLOS ONE review pooled fourteen studies and came out net negative, closing by saying the evidence could not decide the question. It was a review of binaural beats, which is the position monaural audio is usually in: adjacent to the thing being studied.

None of this supports treating any condition.

Every monaural preset is built the same way: a 200 Hz tone, a second tone at 200 Hz plus the beat rate, both held at exactly the same level so the null is complete and the swell reaches full depth. Rates run from 1 Hz to 50 Hz, covering all five bands.

Because the beat is acoustic, none of them need headphones, which is the whole point of the category. Start one on a bedside speaker, put rain or singing bowls over it in the 7-layer mixer, and the swell holds underneath. Save the stack and it reopens as a single mix.

A speaker on the bedside table, playing two tones 10 Hz apart, is the whole setup. Nothing routed, nothing on your head.

Both tones do have to be audible on whatever is playing them. The envelope survives any playback chain, including Bluetooth, compression and a mono downmix, but a small speaker that produces nothing below 150 Hz will not give you a beat built from 60 and 70 Hz, because it never delivers the pair in the first place. If you are generating your own tones, somewhere between 150 and 500 Hz keeps a small speaker honest. The 200 Hz above sits in that window: low enough to stay untiring across an hour, high enough to clear the bass a bedside speaker cannot reach.

Keep the two tones at the same level. Equal levels mean the dips reach silence and the swell is deep. If one tone is louder the envelope no longer reaches zero and the beat softens to a mild wobble. Ten hertz apart, no real speaker will unbalance them meaningfully. A 40 Hz gap on hardware with a lumpy response is where it can drift.

The room does some unbalancing for you. Reflections do not treat 200 and 210 Hz identically, so the two tones are not equally loud at every point in a bedroom, and where their levels differ the cancellation is incomplete and the swell feels shallower. Move the speaker off the wall, or your head fifteen centimetres along the pillow, and the depth changes. It costs nothing to test, and it is the only part of monaural playback you can improve by shifting furniture.

Layering costs you something specific here. The swell is the only sign the file is running, so rain or brown noise over the top takes away your one way to check on it. At 1 or 2 Hz that is a fair trade, since there was barely anything to hear to begin with. Keep the whole stack quiet either way; nothing here rewards volume.

Give it an end time. Half an hour covers the wind-down, and the sleep timer stops it for you.