Past the presets: eight tools for building your own.

Not beginners. If you have used binaural beats fewer than a dozen times, the presets are genuinely the better tool — they encode reasonable defaults and remove decisions you have no basis for making yet.

The Lab becomes useful at the point where you have noticed something the presets do not do. You want a session that starts alert and ends asleep. You want to know whether a square wave feels different from a sine. You want noise underneath without opening a second app. Those are the problems these tools exist for, and if you do not have one of them yet, you are not missing anything.

One honest caveat before the list: most of what you can build here has not been studied. The evidence base covers plain binaural beats at a handful of frequencies, and thins out fast beyond that. Building a bespoke session is legitimate experimentation, not the application of a protocol somebody validated.

Set any carrier and any beat frequency, and watch the waveform in real time.

The oscilloscope is not decoration. Seeing two tones drift in and out of phase is the fastest way to understand what a binaural beat actually is — the pulse you perceive is the interference pattern, drawn on screen. People who have read three articles and still not quite got it usually get it in about ten seconds here.

Practically: this is the tool for finding your number. Individual peak alpha frequency varies between roughly 8 and 12 Hz, and the difference between 9 and 11 Hz is not something you can reason your way to. You find it by sitting with each for a week. The free web generator does this part too, so you can start without installing anything.

A sine wave contains one frequency and nothing else. Square, triangle and sawtooth waves add harmonics: whole-number multiples of the fundamental, in different proportions.

This changes the timbre substantially. A square wave at 200 Hz sounds hollow and reedy; a sawtooth sounds bright and buzzy. Whether it changes the effect is a genuinely open question — almost all binaural beats research uses sine tones, so there is no literature to appeal to here.

What is not in doubt is comfort, and that matters more than it sounds. Harmonically rich waveforms are more fatiguing over a long session and more likely to produce the headaches discussed in the troubleshooting guide. If you are running an hour, sine is the safe choice. Reach for the others when you want the session to feel like something rather than disappear.

Build a session that changes frequency over time, on a visual timeline: ten minutes at 6 Hz, stepping down to 4, then to 2, with defined transitions between them.

This is the single most useful tool in the Lab, and the reason is in the 2 Hz page: going straight to a deep sleep frequency while your mind is still busy tends not to land. Walking down mirrors the natural descent through the sleep stages, and most people find it noticeably more effective than a fixed tone. Doing that manually means fiddling with a slider when you are trying to fall asleep, which rather defeats the purpose.

The reverse works too. A morning session that starts in alpha and climbs into beta is gentler than starting at 20 Hz cold.

Keep transitions slow — several minutes rather than several seconds. An abrupt jump is perceptible and jarring, which is the opposite of what you want.

Run binaural beats, an ambient bed and a noise generator at once, with independent levels.

The obvious use is masking: beats alone in a noisy room lose to the room. Putting brown noise underneath raises the floor so the beats are audible without turning them up.

There is a real caveat, and it is recent. A 2025 randomised study found that masking sound measurably affects the entrainment response — carrier tone and background both matter, not just the beat frequency. So layering is not free. If you are still working out whether binaural beats do anything for you, test them clean first and add layers afterwards, or you will not know which part is doing what.

Ten built-in reference points, and the arithmetic done for you: enter a carrier and a target beat, get the two tones.

Modest, and useful mostly as a guard against a specific mistake. People sometimes assume the carrier is the effect — that 528 Hz “is” the frequency doing something. It is not; the beat is the difference between the two tones, and 528 Hz played to both ears produces no beat at all. The calculator makes that concrete: pick 528 Hz as a carrier, pick 6 Hz as a beat, and it hands you 528 and 534. The frequency guide covers the distinction properly.

Isochronic tones pulse a single tone on and off at the target rate, rather than relying on two tones and stereo separation. The practical consequence is large: they work through speakers, and they work for people with asymmetric hearing, for whom binaural beats may do nothing at all.

The “true sinusoidal AM” detail is worth explaining, because it is not marketing. Many isochronic tracks gate the tone abruptly — hard on, hard off. Sharp edges create audible clicks and add high-frequency harmonics that were not part of the intended signal, which is both unpleasant and imprecise. Sinusoidal amplitude modulation shapes the envelope smoothly instead, so the pulse rate is clean and there are no artefacts.

Reach for this when headphones are impractical — a shared room, a car (not while driving), or a session you want to run on a speaker. The full comparison covers which suits which situation.

Cross-frequency coupling is a real and well-studied phenomenon: slow rhythms modulate the amplitude of faster ones, and theta–gamma coupling in particular is associated in the literature with memory encoding. The theoretical picture is that individual items are held in gamma cycles nested inside a theta cycle.

This tool nests a gamma beat inside a theta cycle to mirror that structure.

Be careful about what that does and does not imply. The coupling research is about rhythms the brain generates during memory tasks. Whether reproducing that structure in audio induces the same functional state is a reasonable hypothesis and not an established finding. This is the most speculative tool in the Lab, and it is worth using with that in mind rather than as a memory technique with evidence behind it. If you want something better supported for memory, spaced retrieval practice beats every audio intervention ever tested.

Alternates the sound between left and right at an adjustable rate.

The important thing about this tool is what it is not. Bilateral stimulation is one component of EMDR, an eight-phase psychotherapy for PTSD; an audio track that pans left to right is not that therapy, and using one to deliberately revisit traumatic material alone is not advisable. We wrote a whole page on why, because the distinction gets blurred constantly and the consequences of blurring it are not trivial.

What it is genuinely good for: an attention anchor, because slow predictable movement is easy to follow and hard to tune out, and a wind-down cue at around 0.5–1 Hz. If a therapist has suggested bilateral audio for between-session grounding, follow their guidance rather than a website.

Eight tools is more than anyone needs at once, and trying them all in an evening is the fastest way to learn nothing. A reasonable order:

1. Custom frequency, for a fortnight, to find the beat frequency that suits you. Everything else builds on knowing that.
2. Sweep designer, once you have a number, to build one session you will actually repeat — a sleep descent is the usual first one.
3. Mixer, if your environment is noisy. Not before, or you will not know what is working.
4. Isochronic, if headphones are a problem, or if binaural beats have never done anything for you.

The waveform selector, calculator, coupling and bilateral modes are worth exploring once the above is familiar. None of them is a shortcut.

Change one thing at a time and give it a week. The effects here are modest and variable, and if you change three variables at once you will have no idea which one mattered — which is exactly the trap the whole field is stuck in.