Binaural Beats for Meditation: What the Science Shows

Put on headphones and play a 400 Hz tone in your left ear and a 410 Hz tone in your right ear. You’ll hear a third sound, a low pulsing at 10 Hz, that exists nowhere in either signal. Your brain is manufacturing it.

This is a binaural beat. It happens deep in the brainstem, in a structure called the medial superior olivary complex (MSO), which normally compares the timing of sounds arriving at each ear to figure out where they’re coming from. When two slightly different frequencies arrive simultaneously, the MSO detects a continuously shifting phase relationship between them and fires at the difference frequency. You hear a beat that has no physical source.

The phenomenon was first described in 1839 by Heinrich Wilhelm Dove, but it was Gerald Oster’s 1973 article in Scientific American that put it on the map for researchers. Oster mapped out the constraints: binaural beats only work when the carrier tones are below about 1,000 Hz (because the auditory neurons that detect phase differences can’t keep up above that), and the beat frequency has to stay below roughly 30 Hz. The sweet spot for carrier tones is around 400 to 500 Hz, where the widest range of beat frequencies is perceivable.

Headphones are non-negotiable. If both tones mix in the air before reaching your ears, you get a monaural beat instead, which is a different phenomenon with different properties. The two tones have to arrive separately.

Robert Monroe and the layered approach

Oster described the basic phenomenon. Robert Monroe spent decades building a system on top of it.

Monroe was a Virginia-based broadcasting executive who became interested in the effects of sound on consciousness in the 1950s. By the 1970s, his research had evolved into what he called Hemi-Sync, short for hemispheric synchronization. His core idea was that specific layered combinations of binaural beats could reliably produce specific states of consciousness.

He was granted three patents for the technology, all now expired and in the public domain. The first (US3884218A, 1975) described amplitude-modulating a carrier sound with signals that matched the waveshape of EEG sleep patterns. The second (US5213562A, 1993) was the significant one: it introduced the concept of using multiple simultaneous carrier pairs, sometimes dozens, to create complex binaural beat environments. It also introduced what Monroe called “Phased Pink Sound,” a stereo pink noise layer whose phase and amplitude rotated in sync with the binaural beats, which his testing showed enhanced the subjective effects by up to 30%.

The third patent (US5356368A, 1994) described a specific configuration called the Septon: five carrier frequencies per channel arranged so that both binaural and monaural beats overlap, producing seven simultaneous beat signals from a single frequency stack. It also specified using averaged EEG waveforms from 40 to 50 individuals in the same brain state as the basis for shaping the beat contours, rather than simple sine waves.

The key insight across all three patents is that Monroe wasn’t working with a single binaural beat. He was building layered auditory environments with three to seven simultaneous carrier pairs, each targeting a different frequency range.

The Focus levels

Monroe’s research program at the Monroe Institute organized these layered configurations into named states called Focus levels. The Institute has never published the exact frequency recipes, but an anonymous researcher measured the actual frequencies on the Gateway Experience tapes, and those measurements have been available in the open-source community since the late 1990s.

What the Monroe Institute calls Focus 10, described as “mind awake, body asleep,” uses four layers: a delta-range beat (1.5 Hz) at a low carrier, plus theta-range beats (4 Hz) at three progressively higher carriers. The target is a theta-dominant brainwave state (4 to 8 Hz) with enough awareness maintained that you don’t simply fall asleep. Meditators have described this boundary state for centuries. Monroe gave it a label and an audio recipe.

Focus 12, which the Monroe Institute describes as “expanded awareness,” retains the entire Focus 10 foundation and adds three more layers on top: alpha-range beats (10 Hz) at mid-range carriers and an additional theta beat at a higher carrier. Seven layers total. The design logic is additive. You don’t replace the lower state. You build on it.

Focus 15 (“no time”) shifts the upper layers to beats around 7 Hz, which happens to match the Schumann resonance, the electromagnetic frequency of Earth’s ionospheric cavity. Whether that correspondence is meaningful or coincidental is an open question. Focus 21 (“the bridge”) is the most dramatically different configuration: the theta foundation persists, but the upper layers jump to beta-range beats around 16 Hz at high carriers. The simultaneous presence of theta and beta frequencies is designed to produce a state Monroe described as dissociated alertness, where the body is deeply relaxed but the mind is unusually active.

The Gateway Experience, Monroe’s home training program, progresses through eight waves of audio exercises that gradually introduce these Focus levels. It starts with simple relaxation and builds toward states that Monroe associated with expanded perception and, eventually, what he framed as contact with non-physical reality. That framing is Monroe’s. What the audio actually does to the brain is a separate question, and the research on that is worth looking at honestly.

What the science says (and doesn’t say)

The most common claim about binaural beats is that they “entrain” your brainwaves, meaning they cause your brain’s dominant electrical frequency to shift to match the beat frequency. This is the claim you’ll find on most product pages and YouTube descriptions. It’s also the claim the research has struggled to confirm.

The most thorough review of the evidence was published in PLOS ONE in 2023 by Ingendoh and colleagues. They examined 14 controlled EEG studies that met strict inclusion criteria. Five supported the entrainment hypothesis. Eight reported contradictory results. One was mixed. The strongest negative results came from Lopez-Caballero and Escera (2017), who tested five different beat frequencies and found zero EEG effects at any of them, and Goodin and colleagues (2012), who tested 31 participants and found no entrainment.

A February 2025 study in Scientific Reports (n=80) is the most methodologically careful to date. It used a factorial design with four randomized parameters: beat frequency, carrier frequency, background noise, and onset timing. The results were mixed in an instructive way. Gamma-frequency binaural beats did produce measurable EEG entrainment, but the effect was significantly weakened when white noise was present. For beta-frequency beats, the evidence was weaker.

There’s a finding from 2002 that nobody talks about much. Stone and colleagues found that EEG changes associated with Hemi-Sync disappeared when electromagnetic headphones were swapped for air-conduction headphones. The entrainment effect may have been partly electromagnetic rather than purely acoustic. That single study hasn’t been replicated, but it hasn’t been refuted either.

So direct brainwave entrainment is, at best, inconsistently supported. If binaural beats worked the way most marketing claims suggest, the EEG evidence would be overwhelming by now. It isn’t.

What actually is happening

The more interesting finding is one that gets much less attention. Several studies that found no entrainment did find something else: changes in functional connectivity between brain regions.

Gao and colleagues (2014) found no evidence that brainwaves shifted to match the beat frequency. But they did detect significant changes in how different brain areas communicated with each other, measured through phase-locking values and cross-mutual information between regions. Orozco Perez and colleagues (2020, eNeuro) found that monaural beats were actually better at raw entrainment, but binaural beats produced unique connectivity patterns that monaural beats did not.

This distinction matters. Monroe called his technology “hemispheric synchronization.” The connectivity research suggests he may have been more right than wrong about what was happening, just wrong about the mechanism. The effect isn’t your brain locking onto a frequency. It’s your brain being forced into an unusual pattern of inter-hemispheric cooperation because it has to process a phantom signal that exists in neither ear.

For meditation practice, this may actually be more relevant than entrainment would be. The states that experienced meditators describe are characterized by unusual patterns of awareness, not by a specific brainwave frequency. Changes in how brain regions communicate with each other are closer to what those states look like on a scanner than a simple frequency shift would be.

The behavioral evidence is stronger

While the EEG entrainment evidence is contested, the evidence for behavioral effects is considerably more solid.

Garcia-Argibay and colleagues published a meta-analysis across 22 studies in 2019 and found a medium effect size (Hedges’ g = 0.45) for binaural beats on memory, attention, anxiety, and pain. That’s a meaningful effect, comparable to many accepted therapeutic interventions. A 2025 perioperative meta-analysis across 14 trials and over a thousand patients found significant anxiety reduction even compared to non-binaural-beat audio.

There’s something puzzling buried in this data, though. The 2025 factorial study found that gamma beats with white noise background killed the EEG entrainment signature. The brain wasn’t frequency-following anymore. But attention performance still improved. The behavioral benefit persisted in the absence of measurable entrainment. Garcia-Argibay’s meta-regression found the same pattern from a different angle: session duration and beat frequency selection were the strongest predictors of effect size, not whether pink noise was present, not carrier frequency precision.

Whatever binaural beats are doing, it seems to work through the total auditory experience rather than through a single identifiable mechanism. The honest framing is that binaural beats create an auditory environment that facilitates certain cognitive and emotional states, and the research supports that framing more strongly than the narrower claim of brainwave entrainment.

How Umbral uses this

Umbral’s audio system draws from Monroe’s multi-layer architecture and the publicly available frequency research. Each session type uses a different configuration designed for its specific purpose.

Grounding sessions use a deep theta configuration: four layers with beats in the 1.5 to 4 Hz range, plus a pink noise bed. The goal is a calm, settled state that supports body awareness and the feeling of being anchored.

Expanded-awareness sessions use a more complex seven-layer configuration that combines a theta foundation with alpha-range beats at higher carriers. This reflects the research on expanded-awareness states, where theta and alpha frequencies coexist.

Some sessions transition between configurations mid-session, crossfading from a theta-dominant environment into a more complex one as the practice progresses from body scanning to active awareness work.

Umbral is not affiliated with the Monroe Institute. We use descriptive terms like “deep theta” and “expanded awareness” rather than Monroe’s Focus level terminology, because what matters for your practice is the state, not the brand name.

The audio is generated from mathematical specifications rather than played from pre-recorded files. Each configuration defines the carrier frequencies, beat frequencies, number of layers, and pink noise levels. This means new configurations can be tuned to different practice states without recording anything new.

A few design decisions worth noting. Pink noise is present in every configuration, but at moderate levels (25 to 30% of the mix). The 2025 research confirmed that background noise degrades measurable EEG entrainment, but behavioral effects persist regardless of noise. We keep the noise because it makes the experience more comfortable and sustainable for longer sessions, and session duration is actually the strongest predictor of effect. We also keep the binaural tones clearly audible above the noise bed, since the 2023 review noted that every study embedding beats fully within pink noise found no entrainment at all.

We don’t mask the tones. You can hear them. Some people notice them more than others. Either way, the audio is designed to create a sustained environment, not to produce a dramatic perceptual effect you’re supposed to “feel.” Think of it as acoustic architecture for the practice state.

Using binaural audio in practice

Stereo headphones are required. Over-ear or in-ear monitors with good isolation work best. Bone conduction headphones work but produce subtler effects. Open-back headphones are acceptable at lower volumes. Speakers will not produce binaural beats at all, because the tones mix in the air and become monaural.

Keep the volume moderate. Research protocols typically target 50 to 60 dB SPL, which is roughly 40 to 50% of your device volume. The developer of SBaGen, one of the oldest open-source binaural beat generators, noted that playing them louder “didn’t improve the effect at all, it just gave me a headache.” If the tones are clearly audible and the experience is comfortable, the volume is right.

Sessions need to be at least eight minutes long for measurable effects. The Garcia-Argibay meta-analysis identified session duration as the single strongest moderator of effect size. This is one of the reasons Umbral’s shortest sessions are five minutes for grounding (where the audio is secondary to the breathing work) and ten minutes for sessions where the audio environment is central to the practice.

Don’t expect to feel the beats as a distinct sensation. Some people perceive a gentle pulsing. Most people don’t notice anything specific after the first minute or two, and that’s fine. The audio creates a background condition for the practice. The practice itself is what you’re doing with your attention while the audio is running. The binaural environment supports the work. It isn’t the work itself. Your attention and your willingness to sit with whatever arises carry the practice. The audio just makes the room quieter inside your head.

Umbral is not affiliated with the Monroe Institute. Focus level terminology and the Gateway Experience are the Monroe Institute’s work, referenced here for educational context.

---

Binaural audio is one part of how Umbral structures daily practice. If you’re wondering where to begin, the grounding practice is the foundation everything else builds on. For the broader context on why grounding matters for sleep paralysis, the sleep paralysis guide covers the neuroscience and the cross-tradition convergence.