BEAT 10.0 HzALPHACARRIER 200 HzMODE BinauralFADES IN 15:00
Signal details
DELTATHETAALPHABETAGAMMA
Sound controls
Session planner
A little space for right now
Choose your preferences, review a suggestion, then begin. Nothing plays until you press Begin this session.
01 Make it yours
Your intention
Time available
Your sound
Before you listen
Listen at a comfortable level. Do not use while driving or when your surroundings need attention.
How you feelmixed · steady
Activation
very settledoverloaded
Energy
drowsyrestless
Room & optional sensorsno hidden sensing
These facts change presentation and safety rules, not a diagnosis. Choose them directly, or ask for a brief local sensor hint and confirm it yourself. No recording, movement trace or heart-rate history is retained.
What is the room likemanual listening context
Can the session be protectedinterruption context
Optional local checksnothing starts automatically
Room activity
Samples microphone level for five seconds. It keeps only a rough level summary, never audio.
Not sampled.
Movement hint
Samples device motion for five seconds. It can suggest still or moving, never driving.
Not sampled.
Heart-rate monitor
Reads a standard Bluetooth monitor. BPM is shown but does not select a session.
Not connected.
Microphone level is digital dBFS, not calibrated sound pressure. Motion depends on where the device rests. Bluetooth support varies by browser. These are suggestions with explicit limits, not validated measurements of stress, attention or consciousness.
Saved listening preferencesNot applied
Motor tempo, binaural release and gap threshold remain observed only. They do not select a session or replace its target rhythm. Review Your Ears.
Suggestions update as you choose. Optional sensors stay off until you ask.
How suggestions work
The planner uses your choices, device clock and records stored in this browser. Unknown context stays unknown. Nothing is sampled without an explicit action and permission. BINAURATUNE does not use your camera, location or private messages here.
This is an explainable recommendation system, not a diagnosis or a brain-state detector. A personal history changes the ranking only after enough matched blinded observations exist. Repetition without an outcome is treated as familiarity, not proof.
Compare personalizationOptional private study
Prove the fitadaptive vs static
Personalization should earn its name
At eligible moments, BINAURATUNE can conceal whether your five-minute session came from the full context policy or from a matched static policy that ignores activation, energy, room, interruption risk, history and clock. Goal, presentation, playback route, sensitivity, safety situation and dose remain matched. Record one before-and-after rating, then the policy is revealed.
Nothing is randomized when both policies would deliver the same effective session. That is not a failed test; it means the adaptation made no decision worth testing. The comparison stays on this device and makes no medical inference.
Instrument guide
Make it your sound.
Start simple. Shape one thing at a time.
01 / SET UP
Check your listening setup
Use stereo headphones for binaural sound. Start at a comfortable, low volume.
These links only open tools. They do not start sound or change your settings.
Help me choose a starting point
Choose a goal to see a suggestion. Nothing changes until you choose an action.
Control reference
Open what you need
What influences what
The controls work together; they are not one long chain of effects.
Method + rhythm + carrier
Define the main sound. Timbre, shape and method-specific controls refine that route.
Hymnal or woven protocol
Uses its own tone layers instead of the ordinary method layer. Resting controls are marked in the console.
Atmosphere + optional layers
Join the active tone source. Breath tide, when enabled, modulates their shared sound bed.
Output
The combined bed passes through volume and narration ducking, dry/wet space, compression, ear-balance filters and the output ceiling.
That is the instrument route, not every imported recording or narrator. Signal diagnostics provides separate output tests.
Rhythm & carrier
Beat sets the requested rhythm from 0.5 to 100 Hz. Carrier sets the tone pitch from 60 to 1000 Hz. They are different quantities: a 200 Hz carrier with a 10 Hz binaural difference uses 195 Hz in one ear and 205 Hz in the other on the basic, unmodified route.
Delta, theta, alpha, beta and gamma are labels on the dial, not measurements of your brain. Choosing a band does not establish a psychological or health benefit.
Drive rate, where available, divides the requested rate. A 40 Hz target at one-third drive produces about 13.33 cycles per second in the sound. It does not demonstrate a 40 Hz brain response.
Blend depths work on the Blend method. Binaural depth scales the ear-to-ear difference; AM sets pulse depth; Orbit sets stereo movement; F-mod sets frequency wobble. Zero removes that component.
For example, a 10 Hz target with binaural depth at 50% produces a 5 Hz difference on the basic Blend route. Adding Orbit introduces channel mixing; it is no longer an isolated two-tone binaural signal.
Comb levels adjust individual carriers in Comb, not the Blend components. The console disables controls that the current route does not use. Read the nearby explanation rather than expecting every slider to affect every method.
Hymnal and woven protocols own their tone layers. Turn music off or end the woven protocol before returning to the ordinary method. Full reveals optional controls; it does not activate them.
Timbre changes the tone texture. Burst uses filtered noise. The Phantom and Hymnal have their own source design, so ordinary timbre controls rest there.
Pulse shape changes a compatible amplitude gate. Isochronic subharmonic drive uses its own gate, so shape choices rest until you return to Direct.
Atmosphere adds noise or a synthesized environment alongside the tones. Choose a type before adjusting its level or enabling noise pulsing. Space mixes in reverberation; it is unavailable on the ordinary Phantom route.
Hymnal replaces the method layer with generative music and its own sound controls. The beat dial still changes its beating rates. It is separate from recorded or AI-generated music in the Music workspace.
Drift varies the requested rhythm over time. Harmonic lattice adds a quieter octave layer on compatible routes. Comfort adjusts supported sound textures; it is not a loudness or hearing-safety guarantee.
Breath tide adds a slow level swell. Breath-carried depth needs a compatible Isochronic or Blend amplitude gate. Spindle loom adds sigma-rate sound packets, not measured sleep spindles.
Voice guide supplies spoken cues for supported practices. It is separate from carrier timbre and recorded narration. Photic is an optional flashing visual, not required for audio; read its warning before enabling it.
Presets load a starting configuration. Journeys change settings over a timed arc; review the arc and controls before starting. They are not wake-up alarms.
Session planner builds a session around your choices. Listening lab contains 46 activities, each with its own purpose and limits. A preference or tapping result is not a brain-state measurement.
Save an instrument setup in Presets for reuse. Saving settings is different from exporting an audio file.
Start at a low, comfortable level. Adjust one feature at a time and pause when needed. More intensity or time is not a promise of more benefit.
For binaural listening, keep channels separate. Device mono audio, spatial processing and other playback effects can change the signal after it leaves this app. The listening check cannot certify your entire physical playback chain.
Use the timer if you want an automatic fade. Save a preferred configuration instead of trying to reproduce it from memory.
Trials compares a protocol with a control using a planned sequence and ratings. Decide what outcome matters before starting, and follow the setup and listening checks.
Different conditions can still sound different. Guessing, expectations, missed sessions and ordinary day-to-day variation affect interpretation. A short run may remain inconclusive.
Personal ratings are not EEG, diagnosis or proof that a sound caused a change. Signal diagnostics answer a separate question: what audio was generated.
Pause if listening feels unpleasant. Do not use attention-demanding or sleep-oriented sessions while driving or operating equipment. The app does not diagnose or treat a condition.
A correct sound frequency does not establish brainwave entrainment, healing or a cognitive benefit. Experimental methods are ideas to investigate, not promises of latent abilities.
Choose a method. Adjust the rhythm, tone and space.
Selected methodBinauralSeparate tones per earChange +
Choose a method. Open About below to see how its controls work.
About this method+
headphones required
Beat · target rhythm
Sound rhythm, not a measured brainwave rate. Arrow keys adjust by 0.1 Hz.
Enter exact Hz
Carrier · the audible tone
The audible pitch. Choose what feels comfortable.
Enter exact HzFrequency quick picksWhy these pitches?
No carrier is proven best for each brainwave band. These are listening starting points, not prescriptions. Only the carrier changes; your rhythm, method and volume setting stay the same. Perceived loudness may change. Use a comfortable level and isolated stereo headphones for binaural listening.
340 Hz comparison: a 2025 study tested 16 and 40 Hz. Its attention finding favored 40 Hz with a 340 Hz carrier; it needs replication. 250 Hz sleep study: used a fixed left tone, not BINAURATUNE's centered pair, and did not compare carriers.
1,000 Hz AM reference: studied sinusoidal amplitude modulation at 12, 39 and 88 Hz. Other pulse shapes and rates are not that protocol. BINAURATUNE's SAM method is spatial modulation, not the sinusoidal AM abbreviation used in that paper.
Traditional pitches
Commonly called solfeggio frequencies. No special healing effects have been established for these numbers.
Volume
Session timer
The timer ends with a gentle ten-second fade.
Compare the methodA / B listening
Hear the method
Compare the static carrier with your selected method. Both previews are loudness-balanced and peak-limited.
Ready. Headphones are required to judge stereo methods.
Gamma burst rate
The fast pulse rate inside the slower envelope. This is a sound setting, not a brain measurement.
Blend mixer · Blend method onlyactive
Binaural detune
A-mod pulse
Stereo orbit
F-mod wobble
Four techniques on one carrier, mixed by depth rather than chosen one at a time. Every active Blend carrier layer follows all four sliders.
Sculpt the carrier mix with these sliders. Purist uses the first four as widely spaced binaural pairs. Phase changes how the layers combine.
Drive rate
Divide the sound's repetition rate without changing the target dial.
What the research supports
In a 23-person MEG study, 20 and 13.3 Hz tone trains elicited a 40 Hz response, but direct 40 Hz stimulation produced the largest response. Comfort was not measured. This does not establish equivalent effects at arbitrary targets or for this app's Flux and pulse envelopes.
For Isochronic, divided drive selects an asymmetric envelope so even harmonics can occur; Flux instead changes its brightness-modulation rate. Harmonics in rendered sound are not a measurement of neural entrainment. Compare how the settings feel without assuming that slower, stronger or more complex is better.
Depth
Phantom · what your ear builds
An experimental distortion-product listening test. Perception varies, and your playback hardware may also add distortion. Hearing nothing is normal. Never raise the volume chasing an effect.
Pulse shape
Tone & musicSettings
Voice · the carrier's character
Choose the carrier's texture. These are synthesized sounds, not narration voices.
Music · the Hymnal
Add generative chords. The beat dial changes their tone-pair spacing; a custom Hymnal keeps its relative spacing ratios.
Direct the Hymnaldusk palette
Shape the score without changing the selected beat. Brightness responds now. Motion, density and width enter on the next unscheduled phrase, usually within thirty-five seconds.
Protocol lockArt direction can change pacing, filtering, voicing and stereo spread. It cannot change the target beat, within-voice frequency spacing or trial condition.
No Long Listen for this palette yet.
earned personalizationYour Listening Signature
collecting
Complete a 20-minute Long Listen on three different dates to begin a repeatability check.
Motionunknown
Brightnessunknown
Densityunknown
Widthunknown
This summarizes repeatable listening preference, not treatment response, brain state or entrainment.
Atmosphere & spaceSettings
Atmosphere
Choose an atmosphere above to use its level and pulse controls. Your settings are kept while it is off.
A background texture under the tones. Start quietly and adjust to taste.
No personal soundscape fit yet.
Space · generated hall
A dark, synthesized hall. A little space makes long sessions far less clinical.
Optional layersSettings
Optional layersexperimental
These change the sound or presentation. Additional layers do not guarantee better outcomes.
Flashing light can trigger seizures in photosensitive epilepsy. Confirm you are not at risk:
Save & exportWAV or settings
Settings save automatically · Space plays/pauses · ← → fine-tune the beat
How this sound is madeSignal diagram
The mechanics · what this method actually doesbinaural
Each ear gets a slightly different pitch; the wobble between them exists only inside your auditory brainstem.
Cymatics
The Chladni Window
A modeled pattern, with a separate view of the instrument's digital signal.
PATTERN PITCH 200 HzMODE (3,1)
Visual motion never changes the sound.
The geometry is illustrative, not a physical plate simulation or a brain-state measurement.
What the picture and readings mean
Pattern pitch follows the carrier setting, even when music or a woven protocol owns the sound. Tide drifts between neighboring pattern modes. With motion enabled, the sampled level and spectral flatness also shape the drawing.
Still view freezes animation, not audio or the readings. Changing a pattern setting can redraw the still image. Reduced-motion preferences start this view still; you can explicitly enable motion.
Live readings come from the instrument's existing analyser after its output ceiling. They update while this view is visible. Stopped, suspended or unavailable audio is not presented as a current measurement.
Approx. spectral peak estimates the strongest non-DC spectrum component, not the carrier or binaural difference. Closely spaced tones, noise and channel cancellation can make it ambiguous. RMS and sample peak are digital dBFS, not sound pressure at your ears. Values below the display floor are shown as such.
This view does not monitor every narration or imported recording. It cannot certify channel separation, entrainment, hearing safety or therapeutic effects. Use Signal diagnostics for dedicated output tests.
States
Find your starting point.
14 starting sounds. Play one, or customize it in the Console without starting audio.
Your saved setups Save, load and recover
Save this instrument's settings under a name. Recordings and mastered audio are separate.
Download the original, then reselect that file. Skipped and invalid records stay in this backup, not the replacement.
Preview only. No sound starts. Review normalized controls before saving.
Save is confirmed by reading the stored copy back. Keep a separate export: browser data can still be cleared or evicted. New saves use a versioned library; older builds cannot edit it. Close older BINAURATUNE tabs before editing saved setups.
Preset names describe listening ideas, not guaranteed outcomes.
No matching sounds. Try another search or clear the filters.
Journeys
Let your sound evolve.
Browse 13 evolving sessions. Choose a duration, inspect the path, then press Start.
How pausing and editing work
Pause holds your place in Journeys and the Ladder. Their progress follows the audio clock, not time with this page open. If sequence updates are delayed by more than two seconds, playback pauses at the last applied step. Keep the page active for the smoothest timing. A playback clock is not proof of what you heard.
Volume changes keep the Journey running and preserve its fade. Editing the sound, including its method, carrier, atmosphere or Hymnal arrangement, ends the script and returns you to manual control. That interrupted sequence does not earn an endpoint record.
Journey history and recovery
Checking saved Journey history.
History keeps up to 120 endpoint records on this browser origin. Resolve an unconfirmed completion before starting another Journey. Recovery drafts remain in this tab only, so export them before reloading. Backups contain your private activity. Clearing or replacing history does not authorize replay of an older draft. Imported SBaGen sequences do not write this history.
Close older BINAURATUNE tabs before using this build. They do not understand the new Journey history format.
A download request does not confirm that your browser saved a file.
Rates describe the planned sound, not a measured brain state.
No matching journeys. Try another search or clear the filters.
Affirmations
Make the words yours.
Write a few lines, choose how they sound, then bring them into your session.
01 / Your words
Up to 40 lines. Edits are a draft until saved.
One line per affirmation. Save the lines you want in rotation. Playback uses the voice source you choose.
Writing and privacy
Use words you would actually say. Saved text stays in this browser. A network browser voice may send text to its speech provider; a prepared AI voice sends saved lines to your selected endpoint only through the preparation flow.
02 / Choose a voice
Device voice and tuning Selected
Pace
Pitch
Pace and pitch affect device speech only. Preview to judge comfort.
Your recordings Not selected
Audio clips
Record or import short clips. Stored in this browser, with a shared 2.5 MB limit. Importing does not upload a file.
Natural local voice pack
Voicebox prepares from the shared service settings in The Sitting. Native takes keep their own engine and preset. Files stay in this tab; back up wanted takes. Every new or restored pack requires listening review.
Uses the Voicebox address, model and delivery direction selected in The Sitting.
Connect Voicebox, choose a profile, then prepare the saved lines you want in rotation.
Complete listening review
A measured file can still contain a skipped word, strange pronunciation, or damaged ending. Play every exact returned WAV from its beginning to its natural end. Seeking or changing speed invalidates that pass. Muting or zero volume also invalidates it. The pack cannot enter Affirmations or Songsmith until you approve what you actually heard.
Approval unlocks only after every current WAV reaches its natural end at normal speed. Review evidence stays in this tab.
03 / In your session
Volume
When they arrive
Uses the instrument's 10-second cycle, not a breathing sensor. There is no need to match it if that feels uncomfortable.
Off. Enable this, then play the instrument.
Instrument Pause stops automatic lines. Stop affirmation voice ends a preview. Guided sittings and listening comparisons take priority.
Optional breathing practice 4 seconds in / 6 seconds out
A separate paced-breathing runner. This is a timing cue, not a measurement of your breathing or heart-rate variability. Use a comfortable pace and return to normal breathing if uncomfortable.
No completed sessions yet.
Use these as personal reminders, not a promise of subconscious reprogramming or a clinical treatment.
Listening setup
Make listening feel balanced
Check your headphones, adjust left and right, then explore only the listening tests that interest you.
01 · Playback
Start with your headphones
Keep the volume low. This check asks what you hear; it cannot detect your actual audio output.
No playback check saved.
02 · Left and right balance
Set a low device volume, then press a frequency. Adjust its slider until left and right feel equally loud. Stop if uncomfortable.
Checking saved balance.
Balance save & recovery
Unsaved adjustments still work in this page. Export them before leaving. Loading the saved copy discards page adjustments, never the saved original.
These private JSON downloads are not automatic restore packages. A changed or unreadable saved copy is never automatically replaced.
How balance works
Full uses 100% of your chosen offsets; Balanced uses 60%. They adjust the instrument's five-band stereo EQ, not every voice or imported recording. Test tones use the full slider offset so you can compare each band. Overlapping filters and downstream processing mean these settings are not calibrated hearing measurements.
Settings stay in this browser. Recheck after changing headphones or audio processing. This is a comfort adjustment, not a hearing test or a diagnosis.
03 · Explore your listening
Each optional activity answers a different question. None is required to use the instrument.
These activities describe your choices and responses on this setup. They do not measure hearing health, brainwaves or a state of consciousness.
The Sitting
A moment for you.
Choose your practice, time and sound. Make space for yourself.
Opening your practice library…
Your session
Build your session
Choose a practice or write your own words. Your voice and background come next.
Optional and stored on this device. Track ease, mood, attention and difficult reactions before and after a session. Changes do not prove that meditation or sound caused them.
Your delayed follow-up is readyReturn to the same three scales and record what lasted.
Narrator
For a more natural narrator, choose a local AI voice or a human recording. Device voices are a basic option, not our quality recommendation. Your saved narrator choice is kept.
Selection only. Choose a narrator, then prepare and review.
Choose a narrator, then prepare and review. Opening-only narration is followed by quiet practice.
Drafts survive only while this page stays open. Exports can contain private reflections. These controls cover sitting completions and their notes, not separate path or check-in records.
Check-in journal and recovery
Check-in saves are checked locally.
Drafts are private but survive only in this open page. Export before reloading. Downloads can contain sensitive ratings and reflections. They are recovery copies, not verified outcome reports.
First check that no other tab is still running the sitting. This action cannot detect other active tabs. Confirmation records an interruption with unknown end time and duration, never a completed session.
One line, if you want it
Saved only after confirmation on this device. Nothing is sent. Keep this page open if a save fails.
Start gently. Keep your eyes open if that feels better, and stop if uncomfortable. Safety · Evidence. Voice services receive the text you choose to send.
Make it your own
Voice production
Prepare, compare and review your narration.
Review your words, prepare the selected narration, then listen before beginning.
Shortcuts open the relevant control. They never generate, approve or play audio.
Plan, listen and review. Your manuscript, recordings and permissions stay with each tool.
Cue sheet
Find a voice you enjoy
Voice Casting Room
Listen without source labels. Rate each voice. Reveal your favorites.
1. Prepare a comparison
Choose a passage and voice sources. Preparing does not play audio.
Qualification passage
A short application-specific screen for warmth, pacing, silence, and calm delivery.
Let your shoulders settle. Inhale gently for four counts, and exhale for six. Notice the quiet after the word stillness. When the next sentence arrives, let it come without hurry. You do not need to change this moment.
Which voices are included?
General casting includes up to four unambiguous browser voices: your sitting voice, your affirmation voice, then browser order. This is the automatic selection; Device voice lets you choose an explicit shortlist instead. Casting uses fixed delivery settings, not the Device voice preview sliders. Eligible studios join separately; preset-only casting includes up to eight RMS-matched Voicebox presets. Browser order and metadata do not rank voice quality.
No prepared comparison. Build a set first.
Compare saved AI voice takes
Add a licensed human reference
Record the exact passage above. This in-tab audition joins only after rights and passage review. It never replaces or renames a full session recording.
Exact audition manuscript
No human reference loaded.
Decoded signal facts will appear here. Recording bytes and samples are not stored in casting history.
2. Listen and rate
Play each voice, then move all five ratings. Names stay concealed until the comparison is complete.
not started
Casting audio is not uploaded. Browser speech may use an OS online service outside BINAURATUNE's control. Other studios receive the fixed passage only after connection and consent.
Rate what you heard
For the first four scales, 1 is low and 7 is high. For fatigue, 1 means easy to hear for a long session and 7 means tiring.
3. Choose your narrator
Choose any revealed voice or use the recommendation. Your ratings stay unchanged. Applying a narrator does not start a sitting.
Sources revealed
Three-passage narrator qualification
Complete all three passages with the same voices and studios.
Personal ratings, not clinical effects or population MOS. Reports include studio origin/path and voice ID, not URL credentials or manuscripts. Recast after changing studios.
What this can and cannot establish
Personal casting adapts randomized listening from ITU-T P.808, not its full laboratory protocol. Meditation voice study. Browser PCM is inaccessible. Local PCM checks verify levels and pauses, not voice quality or therapeutic effects. Familiar voices can reveal identities. Studio and voice IDs do not verify a person or unchanged model.
Narration levels adjust the current sitting and later cues, not its music or tone bed. Fixed comparisons stay locked. Preview levels apply on the next play. These are app gain settings, not calibrated loudness.
VibeVoice local studioresearch preview
Research preview: Microsoft VibeVoice Realtime 0.5B on a computer you control. MIT-licensed. Use preset voices, disclose AI speech and test before commercial deployment.
1. Voice and format
Narration coveragechoose before preparing
Choose an opening or complete timed guidance. Your selected sound bed is independent of narration coverage.
Choosing a format does not generate, approve or play audio.
Connection settings
Text stays on this computer only when this address is localhost or 127.0.0.1. BINAURATUNE never connects until you press Connect or Prepare.
Diffusion steps
Narration level
2. Prepare your narration
Prepare the selected narration format, then listen and review before beginning. Nothing plays automatically.
Not connected. The timed browser voice remains available.
3. Listen and review
Audition the prepared opening, or open Complete guidance to review each cue. Listening and approval remain separate.
Your narration review
This shortcut moves focus only. It never plays, prepares or approves audio. Matching and listening checks are not voice-quality or health certification.
Files and technical details
Runtime passport
Connect to inspect the actual local model, hardware and precision. Render once to add measurements from received PCM.
Opening and complete guidance keep separate recordings. Complete guidance requires every cue to be heard and approved, then plays the cues with narration pauses. Your selected sound layer may continue between cues. BINAURATUNE does not clone voices or remove disclosures and watermarks.
Voicebox supports preset and cloned voices across several engines. Its app is MIT-licensed; models have separate licenses. Connect checks the service. Prepare sends text; returned WAV samples are measured. The voice is shared with Affirmations: changing it stops previews and clears prepared audio in both sections. Download wanted clips first, then prepare and review again.
Opening mode creates one continuous introduction. Complete guidance renders every meditation instruction as its own measured clip, then rebuilds the remaining session as real silence.
Connection settings
Text stays on this computer only when this address is localhost or 127.0.0.1. Voicebox must allow this page origin before the browser can connect.
Choose a profile after connecting. Qwen size is ignored by other engines.
Add a preset voice
Browse the connected service's English presets. Adding one saves a profile in Voicebox, not a cloned voice. Model installation and a listening check are still required.
Nothing is sent until you browse. This catalog does not show whether a model is installed or how natural it sounds.
Connect and choose a voice to see which controls apply.
Pronunciation lexicon no aliases
Teach every Voicebox meditation and affirmation pack how a written term should sound. BINAURATUNE keeps the canonical manuscript beside the spoken alias, applies the longest whole phrase first, and never sends SSML or hidden markup. This follows the W3C pronunciation-lexicon alias model, but Voicebox receives plain text, so every result still requires listening review.
No aliases saved. Add only terms you will audition.
This is an orthographic alias, not IPA and not a guarantee of phoneme-level control. Changing an alias invalidates prepared Voicebox audio because the text sent to the model has changed.
Sentence chunk
Chunk crossfade
Narration level
Compare preset voices
2. Prepare your narration
Not connected. Install or start Voicebox, then connect here. The timed browser voice remains available.
3. Listen and review
Listen to your prepared audio. For complete guidance, hear every current cue before confirming your review.
Recorded passage
Spoken wording · pronunciation adjustments
Prepare the current narration to review its words.
Approval unlocks only after each current WAV starts at its beginning and reaches its natural end. Pausing and resuming is fine. Seeking or hearing an older take does not count. Use normal speed without muting or zero volume.
Confirm after listening
Recording measurements
Cue timing and decoded signal facts will appear here.
Choose a prepared cue to inspect its exact manuscript line.
Cue Doctor · repair one line without rebuilding the session
The canonical cue stays unchanged. Any saved spoken alias is shown separately. Direction support depends on the selected Voicebox engine.
No alternate take is waiting.
Paired listening
Blind Take Duel
Hear the current and alternate WAV behind randomized Sample A and Sample B labels. Both receive only downward digital level matching from measured speech-active RMS. The exact returned samples are otherwise unchanged.
Listen through both samples before choosing.
Take identities will be revealed only after a choice.
A one-line preference is useful for editing, not a scientific verdict about a voice model. If you already heard the labeled previews, identity concealment is incomplete. Repeated comparisons are needed before claiming a stable preference.
Voice Fit Memory no comparisons
Completed concealed choices will accumulate here without saving the meditation text or pronunciation notes.
Nothing has been stored.
A direction becomes a repeatable local pattern only after twelve decisive comparisons across at least three UTC dates and a 95 percent Wilson lower bound above chance. That is a conservative BINAURATUNE product rule, not a validated clinical cutoff. It never changes a voice automatically and says nothing about meditation efficacy, brain activity, or consciousness.
Direct labeled alternate preview
Opening this preview reveals which WAV is the alternate. Use the concealed comparison first when label bias matters.
Optional local transcription can flag a likely omission or substitution. It is a fallible second opinion, not proof that a cue is correct.
Files and technical details
Runtime passport
Connect to inspect the actual local Voicebox API, model, hardware path and profile. Render once to add measurements from the exported PCM.
BINAURATUNE requests no personality rewrite and no effects chain. The research-backed canonical manuscript remains unchanged, while every plain-text pronunciation alias is retained beside the exact render text sent to Voicebox. Complete guidance is generated one disclosed cue at a time and every returned clip stays measurable. BINAURATUNE holds returned clips only in this tab, while Voicebox may keep its normal local generation history on your computer. Voice cloning is never automatic, and a cloned profile remains unavailable until the rights acknowledgement is checked.
Bring your own recording or an authorized narrator. Start with your exact script, then listen and review. Files stay in this tab, without upload.
Recording format
1. Record and import
Export your opening script. Record it without music, then add the finished file. The rest of the session is unguided; your sound bed can continue.
Opening manuscript
No recording loaded. Export a brief before recording so the file can stay attached to one exact opening.
2. Listen to the opening
After importing, listen to the entire recording here. Check the words and the handoff to silence.
File details and removal
Decoded signal facts will appear here. BINAURATUNE does not normalize, denoise, trim, transcribe, clone, or upload the recording.
1. Record and import
Export the timed brief. Record each numbered passage as a separate file, then choose all clips together. BINAURATUNE adds the planned pauses between them.
Numbered cue manuscript
No cue pack loaded. Export the timed brief, record each numbered line as its own file, then select all clips together.
2. Listen to every cue
After importing, select each clip and compare it with Expected cue script. Playback never counts as automatic approval.
File details and removal
Cue timing and decoded signal facts will appear here.
Your confirmation is a personal review, not automated speech verification. The overview checks what your session still needs.
A place to begin
Your next practice.
Original meditations and guided learning paths.
Create a personal practice path +
BINAURATUNE original · exact 20-minute edition
The Aperture
A fully authored session with one deliberate arc: establish a breath anchor, widen into open monitoring, include the emotional tone of the moment, then return without breaking the thread. The words and timing remain fixed, so a human narrator, a local voice, and a later revision can all be compared against the same work.
four movements
1 · Anchorone place, one breath
2 · Widensound, body, thought
3 · Includethe tone of this moment
4 · Returnordinary life, intact
20 min · 15 cues · mostly silence
Not loaded. Your narrator choice remains yours, and the sound can still be music, a measured tone, a mastered source, or silence.
Private action handoff
Your next visible action
Write the action before the primer. The exact meditation never speaks or stores these words. After the final bell, one press begins a quiet work block.
Write one action small enough to begin without another decision.
No timed work blocks recorded on this device
Only completed or ended-early timer lengths are saved locally. Your task words remain in this tab and are never included in history.
Your private edition map
AnchorWidenIncludeReturn
Complete the edition and mark any personally felt peaks. Repeated sessions will appear here without becoming a brain or efficacy score.
Why this sequence, and where the evidence ends
A 2025 sequence study found earlier changes in several self-reported mindfulness skills when novices trained focused attention before open monitoring, but only 27 people completed it, assignment was not randomized, and most final skill levels did not clearly differ by order. This single twenty-minute composition is therefore an authored hypothesis, not a reproduction of that eight-week intervention. A 2026 online comparison using semi-random sampling associated selected chills-inducing music with more self-reported chills during meditation; it did not verify physiological chills or establish clinical benefit. BINAURATUNE lets you mark a personally felt peak during this session, but does not claim its procedural score will cause one or that a button press measures physiology.
Unsaved marks remain in this page only. Export them before reloading. An export is not an automatic restore or a backup of your sitting notes. A reset applies only to this path and prevents older loaded sessions from restoring its marks.
Medito inspired the course, next-session, and local progress structure. BINAURATUNE does not copy or stream Medito recordings or scripts. Their app code is AGPL-3.0 and their original content requires a separate license, so every practice here is independently written and remains available offline inside BINAURATUNE.
Guidance personalization
The Space Between
Find out whether your practice is better served by supportive reminders or by more uninterrupted space. BINAURATUNE conceals the labels, preserves the same language core, and changes only cue count and pause structure.
not started
complete pairs0 of 3
qualifying sessions0 of 6
next sessionArm A
current defaultSome practice
Begin when you can give one five-minute sitting your full attention. Each arm must happen on a different date.
Held constant
Practice, voice, speech rate, pitch, language core, five-minute clock and silence-only bed
Changed
Reminder count and the silence between reminders
Primary rating
Would I choose to practice this way again?
Decision rule
Three complete pairs and at least a 0.5-point mean difference on a 1 to 7 scale
No conditions have been revealed. Labels stay hidden until both arms of a pair are rated.
The difference is audible, so this is concealed presentation rather than guaranteed blinding. Voice preference is measured separately in the Voice Casting Room. This tool learns a listening preference, not meditation depth, treatment response, brain state, or consciousness.
The music room
Give your moment a sound.
Choose a palette. Add your ideas. Make it yours.
Use the prompt and style controls below to shape your music. Local AI generation is optional and needs a separately installed model.
Plan the movements, then vary one musical choice at a time.
Score blueprint
Your prompt
Shared by the synth sketch and local service. The sketch reads keywords; local AI receives the text. Installed-model and hosted editors keep separate drafts. Up to 1,000 characters.
Try a prompt idea
Sound & mood
Style
Mood
Pace & length
Tempo
seed the tempo from your own measured spontaneous tempo (Your Own Tempo, in the Lab)
Arrangement
Estimated sketch length depends on tempo.
Ready to sketch. Nothing is playing.
Voice & beat layersAffirmations on · beat off
Uses your Affirmations voice. Spoken lines play live and are not included in the instrumental export.
Voice and timing details
AI narration needs your complete prepared and approved voice pack. No device voice is substituted. Music lowers during speech; busy phrases are skipped, not cut short. Stop also cancels this song's voice.
Turns pad voices into beating pairs at the current instrument rate. A signal choice, not a proven brainwave effect.
Write affirmations for this song
Write me affirmations uses local templates, not an AI model. The other button copies a prompt for an AI you choose. Review the resulting lines in Affirmations.
Local & free · model setup requiredGenerate a richer AI soundtrack
Connect to your local ACE-Step service to begin. No model is bundled into this page.
Local service
Local API addressnot connected
ACE-Step needs a separate local installation. Generate sends your brief, settings and any lyrics to this service.
Model & connection options
API key, if configuredsession only
Generation modelserver reported
Diffusion steps
Only a local service address is supported. Optional tone protocols are mixed separately from generated music. Neither is proof of a psychological benefit.
Voice & lyrics
Vocal plan
Instrumental mode sends no lyrics. Any lyric draft stays in this tab.
Review these exact request lyrics. Drafts stay in this tab and do not change saved affirmations. Short songs may not fit every word.
Arrangement & saved-line coverage
Compact uses the first seven saved lines and the last line; shorter lists repeat a chorus. All lines creates numbered verses without dropping lines. The editor limit is not a provider capacity guarantee. A new generation keeps the current master until a valid replacement is ready.
Last captured request lyrics
A local request snapshot, not proof of server receipt or faithful singing. Editing the next draft does not change it.
Generate your track
Request preview loading. Nothing has been sent.
Your track
Reimagine this track
Start with this track. Adjust your brief, choose the change, then make a variation.
Generate a track first.
Generate a source track to begin.
Passage edits may also change sound outside your selection. Your current track stays until the returned version passes sample checks.
How variations work
Uses your current brief and generation settings, with the source track's duration. Influence requests similarity; it does not measure it. Your model must support the selected task.
One pre-edit original stays in this tab until the next successful generation or Forget. Download versions you want to keep. Stop waiting stops browser checks, not the service's GPU work. Sample checks do not judge musical quality.
What changed?
Compares the original and returned samples at the same positions. No time alignment, normalization or resampling. A difference is not a quality score. Listen to both versions before choosing; exact provider WAVs stay unchanged.
Protocol level & mixed export
Protocol level
Uses the console's current method, target and carrier. This is an acoustic layer, not evidence of a psychological benefit. The original master stays unchanged; mixed export creates a new WAV with the source and protocol settings recorded.
Audio facts & file management
The original WAV download is byte-for-byte what ACE-Step returned. BINAURATUNE measures its actual samples for channels, sample rate, duration, peak, RMS, DC and clipping. Tempo and key are independently estimated from the PCM, with uncertainty and metrical ambiguity shown. An estimate is not ground truth.
Local blind comparison
Composer Duel
AI music or browser composer? Listen to both full excerpts, then choose. Levels are matched by lowering, never boosting, the audio. Equal RMS does not mean equal perceived loudness.
Prepare a pair after the AI master has passed sample checks.
Creator productionSession Mastering Room
Choose music or No music, then add narration and optional console sound.
A new session starts from the beginning.
Music
Music source
No track selected
Choose a file or an AI music result.
File details and ownership
No music source loaded. A selected file is decoded and inspected locally before it can enter a session.
Audio stays in this tab. Original files are not uploaded, denoised, normalized or stretched. Assembly uses the instrument's bounded output chain. Source ownership stays with its rights holder.
Voice
Prepare your meditation's voice, then bring it into this mix.
The key links production records to Voice Proof. It is not required to preview a music bed. Prepared narration keeps its own consent, rights, manuscript, and sample gates. Timed browser speech cannot be placed in a downloadable WAV because browsers do not expose its PCM.
Console layer
The protocol snapshot will be named here. It remains an acoustic layer, not evidence of a mental or medical effect.
Captures the tone method, target rhythm, carrier, timbre, modulation and blend or comb settings. Atmosphere, Hymnal, woven layers, sequences and optional Console layers are not copied. Voice and music have their own controls in this mix.
These are requested settings, not measurements of the audio or your brain. A precise acoustic rhythm does not establish a mental or medical benefit.
Balance
Presets adjust backgrounds, not voice. Reactivate to apply background changes.
Background balance
Uses your selected voice. Adjustable during playback.
Full WAV export supports sessions up to ten minutes and requires a prepared PCM narrator, such as licensed human cues, a licensed human opening, or a reviewed local AI opening. Longer sessions remain playable and can export a complete production record without copying the source audio into that record.
Keep your finished session
Saving in this browser is a separate choice. Keep a downloaded backup too.
Browse freely with one idle capture. Replace it when ready. A pair locks the cue.
Stage two input levels. Your live mix stays unchanged and nothing plays.
Keeps your current mix unchanged. Prepare a music bed and switch the tone layer off first.
What stays fixed?
Voice, timing, music, ducking and chime inputs stay fixed; only music level changes. Shared playback attenuation replaces RMS matching. Compression can still alter the mixed voice. Fingerprints do not prove intelligibility or health benefits.
Variant Aempty
Variant Bempty
Capture to preview. Add a second mix for the same speech; playback matches the quieter RMS.
Listening review · optional
Activate your prepared mix, then capture a preview.
Comparison files and limitations
Editing aid, not a formal test or evidence of health benefits. Labeled listening or remembered capture order limits blinding. Equal RMS is not equal loudness. Audio stays in this tab unless explicitly exported.
Portable authored contentCreator Release Studio
Package format and signature limits
Turn the latest completed Session Mastering Room WAV into one portable creator release. The download is a standard uncompressed ZIP shaped as an RFC 8493 BagIt package: exact audio, a machine-readable release record, SHA-256 manifests, and optional creator-key approval.
A signed release proves that the matching private key approved these exact bytes. It does not prove the creator's legal identity, ownership, independence, clock, or rights declaration. BINAURATUNE uses browser-native P-256 and SHA-256, but this package is not a C2PA Content Credential and has no certificate trust chain or trusted timestamp.
Your signing key
Creator Key Desk
Create or load a creator key to make signed releases. Private key material is held only in this tab and is never placed in BINAURATUNE storage.
Public keys and approval records
Signer Shelf
Pin a public key only after opening and reviewing a valid signed release. A pin recognizes the same key on this device. It does not verify legal identity, ownership, endorsement, revocation status, or C2PA trust.
A key-approved ledger requires a loaded creator authority whose exact public key is already pinned here. Its signature proves approval by that key, not who controls the key in the real world.
No public trust ledger is open.
Loading public signer decisions from this browser.
Approval Record Shelf
After reviewing a valid signed ledger, you may keep its canonical signed content on this device. Every record is signature-checked again when loaded or exported. Records contain public keys and review history, never private keys or audio. They prove matching-key approval, not legal identity, ownership, revocation, trusted time, or C2PA trust.
Loading locally kept approval records.
Compare two approval records
Comparison reverifies both signatures and reports exact A-versus-B differences. A key or release present only on one side is not evidence of addition, removal, chronology, revocation, identity, ownership, or endorsement.
Keep two distinct verified approval records to compare them.
Inspect comparison evidence bundle
One bounded ZIP can carry the unsigned comparison and both canonical signed source records together. Opening it verifies ZIP structure, CRC, every SHA-256 payload digest, both source signatures, both record identities, and a fresh comparison. It never saves or merges any record.
No comparison evidence bundle is open.
Audit a selected record set
Select 2 to 12 kept approval records. BINAURATUNE reverifies every signature and describes exact key presence, public-card variants, local review-clock variants, and reviewed release coverage across the set. It does not infer addition, removal, chronology, revocation, identity, ownership, or endorsement.
Select at least two records to build an order-independent set audit.
Inspect set evidence bundle
Opening this ZIP verifies its structure, CRC, every SHA-256 payload digest, every source signature and record identity, then rebuilds the complete set audit from the bundled sources. It never saves, merges, pins, or orders the records.
Export a complete mastered sitting first, then describe the release and complete every rights gate.
Release identity
Release gatesrequired each time
Open a finished release
Open a creator releaseverify before play
BINAURATUNE checks the ZIP structure, CRC values, BagIt payload list, every SHA-256 digest, the release schema, the WAV's embedded production record, the actual PCM, and any creator-key signature. Nothing plays or enters the private library until you review it and press a separate button.
No creator release is open.
Private offline ownershipSession Library
Opening the private library on this device.
Only saved when you choose. Stays on this device.
Collection
Saved on this device only. Keep a backup before clearing storage.
This opens a finished BINAURATUNE WAV with its embedded production record. A WAV does not include editable voice clips, music sources, saved listening positions or ritual order. An unsigned record cannot prove authorship or ownership.
Bring your session names, saved places and playlist order with your finished WAVs.
Keep playlist order too: use Move a collection. Finished WAVs do not preserve editable voice and music sources.
Storage limits and technical details
Audio files and playlist metadata stay in this browser's IndexedDB, not in an account. Saving here does not upload your audio. Request durable storage to ask the browser for stronger retention, not a guaranteed backup.
The library holds up to twelve sessions, 160 MB per WAV and 512 MB of audio in total, or less if the browser reports insufficient space. Quota readings are approximate.
The normal JSON data export does not include session audio or playlist manifests. Download each session from Collection and use each playlist's More menu to export its manifest before moving browsers or erasing site data.
Playlists
Arrange sessions from your collection into a listening ritual.
More
Your draft
Saved on this device
Choose a name for the editor. Saving is a separate step.
Create a ritual, then add masters from Collection.
Playback and export details
Playlists reference your saved audio. BINAURATUNE does not remix, normalize, duplicate or upload it. Automatic handoff is not guaranteed to be sample-gapless.
Ritual metadata stays beside the audio in this browser's IndexedDB. The ordinary manifest preserves browser matching facts. The Android kit independently checks, measures and hashes every selected WAV, then downloads the unique WAVs with one native order manifest. Neither format contains manuscript text.
Complete production evaluation
The Full Listen
Compare two complete versions of the same meditation, without seeing which is which while you listen.
How comparison works
Compare two complete mastered editions of the same exact manuscript and clock. BINAURATUNE verifies the stored WAV bytes, conceals their identities, attenuates only the louder master to the quieter master's measured whole-file RMS, and requires each file to reach its natural end at normal speed before ratings unlock.
Choose two mastered versions of the same manuscript and duration.
Sample 1not heard
Sample 2not heard
Pausing and resuming is allowed. Seeking, changing playback speed, or changing the review level invalidates only the current pass. A valid earlier complete pass remains valid. Use the same headphones and system settings for both samples.
Rate the completed sample
Move every control. One number cannot stand in for the others, so BINAURATUNE keeps each production dimension separate.
Both complete samples and both rating forms are finished. Which complete production would you choose to use again?
Music and transition notebook
Keep sample-specific observations about flow, voice balance and the final ending.
Store two complete masters of the same manuscript, then choose them above. Nothing starts automatically.
Import and combine review receipts
Independent review ledger
Import text-only receipts from separate listeners, or include the receipt made in this tab. BINAURATUNE validates the schema, complete-playback gates, exact pair of PCM fingerprints and rating ranges, then reports counts and means without declaring a universal winner.
No review receipts are loaded in this tab.
This is a production-quality review, not a formal laboratory panel. The preparer may remember which files were selected, unsigned receipts can be copied or edited, equal RMS is not equal perceived loudness, and preference does not establish meditation efficacy, neural entrainment, therapy, healing, or access to latent abilities.
Tool directory
The right tool, in the right place.
Production tools now live beside the work they support. Choose a destination. Nothing starts automatically.
Audit the final WAV against the intended words. Signal facts come from decoded samples. Word agreement comes from the transcript you paste or explicitly request from local Voicebox Whisper. Neither one is inferred from provider settings.
Load intended words and one WAV. Audio and transcript remain in this tab. A loopback transcription request happens only when you press its button.
Compare voices
Narrator Endurance
Compare two 18 to 25 minute voice-only readings. Listen to each complete sample before the names are revealed.
How the comparison works
Actual WAV samples determine whole-file RMS matching by attenuation only. Both recordings must read the same manuscript.
Decision rule: use the exact same two WAV files and manuscript for every repeat. Collect at least three completed receipts across at least two UTC dates. Stop when one narrator leads the mean score by 0.50 or more and also leads direct preference. If that does not happen after five receipts across at least three dates, stop and record no practical separation.
This is a personal production screen, not a formal speech-quality panel. Equal whole-file RMS is not equal perceived loudness, the person loading the files may remember their identities, and a preferred narrator does not establish meditation efficacy, neural entrainment, therapy, healing, or a change in consciousness.
1. Choose recordings
2. Confirm the manuscript
3. Listen and rate
Sample 1not heard
Sample 2not heard
Pausing and resuming is allowed. Seeking, changing playback speed, or changing the matched review level invalidates only the current pass. Keep the headphones and system volume unchanged across both samples.
Rate the completed sample
Move every control. Fatigue and attention pull are kept separate from pleasant first impressions.
Both complete samples and both rating forms are finished. Which narrator would you choose for another long sitting?
Import comparisons
Before sharing, review narrator labels and imported descriptive notes for private information. Their wording is not verified.
Saved comparisons and recovery
Receipts contain labels and notes. Review them before sharing.
Load two distinct voice-only WAVs lasting 18 to 25 minutes. Files and manuscript stay in this tab. Only measured receipts can be kept locally.
Documented voice permission
Narration Rights Dossier
Connect a manuscript and narrator to your permission document. Scope and consent remain your declarations, not legal verification.
1. Select source files
2. Describe the permission
3. Review and document
What this record does and does not prove
Bind one external agreement, license, consent record, or terms snapshot to the exact manuscript and narrator production key. BINAURATUNE measures the document bytes and records your declared scope. It never reads the document as legal text, uploads it, stores it, or decides whether the permission is legally sufficient.
Design basis: California AB 2602 requires reasonably specific intended uses for covered digital-replica provisions. The U.S. Copyright Office digital replicas report examines written authorization and license scope. C2PA UX guidance distinguishes measured assertions from creator-entered assertions and warns that a valid signature does not by itself establish trusted identity.
This is a production evidence organizer, not legal advice, a consent ceremony, a contract, a rights registry, a trusted timestamp, or verified identity. A document hash proves only which exact bytes were referenced.
Nothing is uploaded or saved. The exported dossier contains hashes and creator-entered declarations, never the evidence document, manuscript text, or source filenames.
Exact-key approval and review state
Rights Authorization Desk
Create an authorization, check signed files, or update their review status. Signatures do not establish legal identity or permission.
Signing key
Only needed for signing. The private key stays in this tab.
Create a signed authorization
Check signed files
Issue a status statement
Use the current authorization and its matching signing key.
Local signed status records
Review signed status files before keeping them on this device. Pair verification and the Narration Passport will then refuse an older active statement when this shelf already contains a newer withdrawal, supersession, or review.
Checking saved records.
Nothing is kept automatically. The shelf is bounded to 64 verified statements and contains public signed status records, never a private key, dossier, evidence document, or manuscript.
How authorization and status checks work
Sign one exact Narration Rights Dossier with a Creator Key, then issue a separate time-bounded status statement for that authorization. Changing the dossier, validity period, review state, or signing key creates different bytes and requires a new signature.
Design basis: W3C Data Integrity separates cryptographic authenticity from the meaning of a claim. W3C Bitstring Status List models status separately from the signed credential. C2PA security guidance distinguishes revocation, compromise, expiry, and trusted timestamps.
This local format is not W3C VC, C2PA, OCSP, a legal identity credential, or a public revocation registry. The shelf detects only newer signed statements that this device has explicitly kept. It cannot prove that no newer withdrawal exists elsewhere or that a signed review time came from a trusted clock.
Private keys stay in this tab and are never saved automatically. Create or load a Creator Key, sign an exact dossier, then issue a status statement whose next review is no more than 90 days away.
Exact-artifact production lineage
Narration Passport
Bind one directed manuscript to its measured narration, repeated endurance choice, rendered Session Master, and complete-production listening evidence. Every stage is checked from exported artifacts. A filename, provider setting, or checkbox cannot substitute for a matching content fingerprint.
Open a production stage
Load the seven text-only artifact sets. Nothing is uploaded, played, generated, or stored automatically.
A qualified passport establishes an internally consistent production chain for these exact files. It does not authenticate people, prove legal ownership, guarantee model consent, certify accessibility, or establish meditation, clinical, neural, or consciousness effects.
ACE-Step structure, BINAURATUNE authorship
Score Blueprint
Design exact musical movements before generation. Each movement gets its own clock, prompt, tempo and production contour, so a long meditation can be generated in reviewable pieces instead of one opaque request.
No model runs here. The blueprint is a deterministic production brief that can feed the browser composer or your local ACE-Step workflow.
Medito course logic, original curriculum
Path Forge
Build a private practice curriculum from BINAURATUNE's independently written techniques. The sequence begins with concrete regulation and attention skills, then widens only when the selected goal and experience level support it.
The path is stored only on this device. Completion records practice use, not improvement, meditation depth, or a brain state.
Long-form voice direction, provider neutral
Cue Director
Turn exact words into a timed cue sheet for a human narrator, VibeVoice, Voicebox, or another consented studio. Spoken text is never rewritten. Direction, speaker assignment, and silence remain separate metadata.
A cue sheet is a production plan, not generated speech. Every rendered take still needs sample checks, transcript comparison, and a complete listen.
Controlled ACE-Step iteration
Variation Lab
Create three generation briefs from one Score Blueprint movement while changing one planned musical dimension. This makes creative iteration interpretable instead of changing prompt, tempo, density, and instrumentation at once.
Design a Score Blueprint first. Generated WAVs may ignore a prompt, so the returned samples and the complete listening experience remain the source of truth.
Participant and analyst
Path study
Compare a guided sequence with access to the same session library. This measures session completion, not meditation efficacy.
Build a Path Forge plan first. No streak, meditation score, mood inference, or consciousness label enters this study.
Checking saved study data. Keep this tab open if a save is not confirmed.
Prepare enrollment and assignment
Build the exact practice path first. Decide eligibility and consent before assignment. Use a nonidentifying roster reference.
Choose the route before enrollment
Load the coordinator public file before committing. Leave it empty only for a local pilot.
1. Commit enrollment
2. Apply the assignment
For a separate coordinator, export the condition-free commitment and receive a signed assignment. A signature verifies bytes and a key, not identity or independent recruitment.
Local assignment is for pilots. It does not provide independent allocation concealment. Open Coordinator for the separate-device workflow.
Assigned sessions
Loading prepares a session. Ending and revealing closes the study early; it is not a pause.
Review completed receipts
Analyze finalized receipts to compare completion. A new import keeps the previous report until its replacement is ready.
Storage and recovery
Study design, evidence and limits
Test a visible sequence against access to the same session library. A shared study code and numbered participant slot create verifiable balanced blocks of four. The endpoint, observation window and planned sample are sealed before Schedule A is assigned. Natural session completion is recorded, while stopping early or running past the window earns no completion.
Why completion, not intention: a 2026 field study of 18,559 Spotify users found that optional meditation goal prompts did not improve population-level meditation frequency. Among people who answered, goal setting was modestly associated with more meditation days, while option order changed goals and opt-out rates. Calm adherence research likewise prespecifies objective app-use windows. Neither source establishes that a visible sequence will outperform a matched library. That is the hypothesis this tool tests.
Why the analysis follows the assignment: permuted-block methodology warns that an analysis should preserve the blocks used to assign participants. BINAURATUNE therefore reconstructs the exact distribution across all six allowed assignments in every block. As the randomization-inference framework makes explicit, the resulting p-value tests the sharp null that presentation changes nobody's completion outcome. It is not the probability that the hypothesis is true and it does not test the weaker claim that the average effect is zero.
Why enrollment now has two steps: CONSORT 2025 separates the decision to enroll from access to the assignment. First commit an eligible participant and externally recorded slot without creating Schedule A. Export that condition-free receipt to the Coordinator Desk when a separate person or device will assign the schedule. Local assignment remains available for pilots, but it does not provide independent allocation concealment.
This is a sealed local analysis charter, not public preregistration or central randomization. Fixed blocks can become predictable. Slot assignment must be recorded before Schedule A is revealed, and local assignment is not an independent central randomization service. The participant reference is hashed and the raw entry is forgotten. A verified signed packet proves only that the holder of the declared study private key signed those exact assignment bytes. It does not authenticate the person's identity, independence, clock, device security, eligibility decisions, recruitment completeness, or seed custody. Use the Coordinator Desk on a separate device controlled by someone outside recruitment when allocation concealment matters. A group result estimates completion behavior for this exact Path Forge plan. It does not establish meditation efficacy or explain why a presentation changed use.
Group study preparation
Flux cohort
Seal the plan before enrollment, then review one completed packet per planned slot. Group direction stays concealed until the full set is present.
Keep the private signing authority off participant devices. Export it before closing this tab. Distribute the public pin separately before enrollment.
Create or open one public charter before enrollment. Nothing is uploaded or stored by this desk.
Create or open the public charter
1. Signing authority
2. Seal the plan
Or open an existing charter
Review completed participant packets
Participant means, not individual taps, are the analysis unit. A complete slot set does not prove unique people, consent or honest follow-up.
Verify a signed full-sample report
A signature binds exact bytes to a key. It does not verify the signer's identity or prove the study's claims.
Study design and interpretation limits
Create a public study charter before enrollment, then validate one completed study-bound Flux packet for every numbered participant slot. Group direction stays hidden until the planned slot set is complete. Participant means, not individual taps, are the unit of analysis.
A new study uses a tab-only P-256 signing authority. Export its public pin through a separate pre-enrollment channel, then export the signed charter. Participants must match both files before sealing a study plan. Legacy signed and SHA-256-only charters remain readable.
This tool can verify exact file bytes, a P-256 key signature, charter compatibility, complete numbered slots, PCM summaries and participant-level calculations. A valid signature proves only that the matching private key approved those exact bytes. It cannot prove who controls the key in the real world, unique people, consent, eligibility, independent recruitment, complete follow-up, honest slot assignment, device conditions, trusted time, or key custody. Unsigned participant packets are not identity evidence. A completed result estimates this contributed slot set under the sealed analysis, not EEG synchronization, consciousness, cognition, treatment, or a general population effect.
Separate coordinator device
Coordinator
Prepare a public study file and issue signed assignments from committed enrollment.
The private file contains the assignment secret and signing key. Keep it off participant devices. This tab is not a backup; export and protect the private file before closing.
Build the exact Path Forge plan before creating a study authority. Nothing is stored after this tab closes.
Create or open the study authority
Build the exact practice path before creating an authority. Use a device controlled outside recruitment. Publish the public key identity before enrollment.
Issue a participant assignment
Open the participant's condition-free enrollment commitment. Issuing an assignment is a separate, explicit action.
Key custody and verification limits
Create a study authority on a device controlled outside participant recruitment. The private file contains the randomization secret and signing key. Keep it off participant devices. The public file can be distributed freely. A coordinator later opens one enrollment commitment, creates the hidden arm from the private seed, and signs the exact assignment packet.
RFC 8785 canonical JSON makes the signed bytes deterministic. Web Cryptography supplies ECDSA P-256 with SHA-256 signing and verification. A valid signature detects changed packet bytes and binds them to one public key. It does not prove who held that key or whether the private file stayed private.
The private authority file is equivalent to the randomization office. Anyone who obtains it can reconstruct assignments and sign valid-looking packets. BINAURATUNE does not upload, escrow, back up, recover, or authenticate it. Use a separate coordinator device, protect the file outside BINAURATUNE, record its custody, and publish the public key identity before enrollment. This tool improves role separation but is not a substitute for an institutional randomization service.
Legacy protocol interoperability
SBaGen Bridge
Inspect a conservative subset of the original SBaGen sequence format, play the translated timing through BINAURATUNE, or export a BINAURATUNE Journey as a readable .sbg file. The receipt separates preserved protocol facts from engine-specific sound.
Sequence tools load when this card comes into view, or when you inspect a source.
This is a strict protocol translator, not an SBaGen emulator and not PCM parity. SBaGen and BINAURATUNE use different gain normalization, noise synthesis, fades, master processing, and output calibration. Unsupported syntax is refused rather than approximated silently.
Open or paste a sequence. Nothing is stored, uploaded, or played until you inspect it and press Play.
About the connected projects
Repository boundary: VibeVoice, Voicebox, Medito and ACE-Step UI remain separate projects with their own licenses and model or content terms. BINAURATUNE calls only consented local services and implements these planning and verification workflows independently.
Study tools
Plan. Coordinate. Review.
Research utilities with explicit preparation and review. Opening a tool does not enroll anyone or assign a study.
Path study
The Lab
Listen. Explore. Compare.
Listening experiments and personal measurements, inside your console. Choose a tool to see its instructions and evidence.
What the evidence labels mean
The label rates the central use of each tool as built. Replicated means the underlying measure or effect has reproduced. Controlled means a closely related human protocol has an active comparison. Exploratory means the human mechanism is indirect or early. Hypothesis means BINAURATUNE's proposed use has not yet been established. These browser tools are not clinical tests.
Help me choose a test
No Lab tools match those filters. Clear one filter or try a broader word.
Open a tool to see its controls and evidence. Sound starts only when you choose its listening action.
Explore 46 tools. Use Trials for repeated comparisons. A listening result is not a brain measurement or proof of benefit.
Trials
Compare your listening
Test a question with repeated sessions. Set a plan, listen, then rate your experience before the reveal.
Start a comparison
Your sound, with a control
Uses the current console settings. Review duration, playback and your plan before anything starts.
PlanChoose what to measure and when to stop.
ListenThe assigned condition stays concealed.
ReflectRecord your guess and ratings, then reveal.
Audible differences can reveal the condition. Personal ratings are not brain measurements or proof of a health benefit.
Practice scheduleOptional cue-linked reminders
The Practice Plan
Choose a time, place and existing routine as your cue. Calendar downloads are optional; BINAURATUNE does not track attendance.
Days
Duration
If the planned time is missed
No plan is stored yet. Everything you add remains on this device.
Evidence boundary: cue-linked plans are supported by implementation-intention research. Multiple schedules, calendar export and recovery windows are practical design choices that BINAURATUNE has not independently shown to improve adherence. Attendance is never tracked and a missed appointment never becomes a failure mark.
Plans & next sessionSealed charters and progress
The first allocation seals your outcome, worthwhile difference, spacing and stopping rule. Confirmatory results stay hidden until the planned target is reached.
Your first personal Trial will seal a charter for its exact protocol. Classroom studies keep their instructor-defined plan.
Evidence Queue
Your next comparison
Suggested from progress and spacing, not interim effects. Opening setup does not allocate a condition or play audio.
Next evidence actionPrepare the first charter
A recommendation will appear from sealed progress, block closure and spacing facts only.
The queue prioritizes completion efficiency, not expected benefit. It never chooses the hidden condition. It receives no rating values, guesses, condition identities, interim contrasts, or future allocation positions. The exact protocol must still pass its normal Lab and delivery checks before allocation.
Why this recommendation?
Design basis: the ACE reporting extension requires adaptive allocation rules and their inputs to be planned and disclosed. FDA adaptive-design guidance explains why response-adaptive randomization needs careful control of bias and operating characteristics. BINAURATUNE therefore keeps the concealed arm schedule fixed and adapts only the order of unfinished protocol work.
Methods & references
Method basis: CENT 2015 asks N-of-1 trials to prespecify the primary outcome, planned measurements and stopping rules. SPENT 2019 asks protocols to state period and washout duration with a rationale when carryover is plausible. BINAURATUNE therefore seals the listener's spacing choice but does not claim that any option is biologically validated. CONSORT 2025 item 14 defines an outcome through its variable, change metric, aggregation and time point. Butcher et al. 2022 asks within-person change trials to define and justify a minimal important change. Lakens, Scheel and Isager explain why the smallest effect of interest should be fixed before results are known.
A local digest detects accidental changes but is not a public registry or tamper-proof timestamp. The composite score, secondary ratings, running journal and cross-protocol summaries remain exploratory. A charter may be retired, but its original plan and early retirement remain in the record.
Optional tests & randomnessStandalone probes and external-data choice
Try a standalone probe. Paired reaction and vigilance tests can also be added during comparison setup.
randomness: browser cryptographic randomness
External supplement off. Browser randomness is sufficient for randomized draws.
Off by default for each page visit. Enabling this contacts ANU and exposes your network address to that service. No trial results or personal text are sent. Supplied bytes are mixed with fresh browser randomness; their claimed quantum origin is not independently verified here.
Running insightsExploratory, not your sealed result
Exploratory running summary across eligible sessions and settings. It can help generate a new question, but only the sealed Trial Charter above defines a confirmatory primary result.
Assignment recordsCompleted and missing outcomes
Every allocated position keeps its completion or missing-outcome record. Conditions and future positions stay concealed while a block is open.
No concealed positions have been allocated yet.
Confirmatory estimates still use complete blocks only. Reporting abandonment does not remove missing-data bias, but it makes that bias inspectable instead of silently treating missing outcomes as if they never existed.
Lab protocol recordsEvidence for each exact recipe
Exact Lab protocols stay separate by their SHA-256 fingerprint. A changed recipe begins a new row rather than borrowing evidence from the old one.
No qualified Lab trials yet. Qualify a fixed protocol from the Lab, then repeat both concealed conditions across several days.
The report uses deterministic randomization and bootstrap draws, adjusts protocol p-values with the Holm method, and uses early-versus-later agreement only as descriptive repeatability. It does not measure EEG, establish causation beyond the randomized contrast, prove a universal effect, or rank protocols that were not directly randomized against one another.
Your journal
No sessions yet. Your evidence starts with the first one.
Lab history & dataObservations, reports, import and export
Explore recorded listening activities. These observations do not establish brain activity or a stable ability.
No observations yet. Run a blinded session or any Lab probe and this sky begins to fill.
Private, unencrypted records only. Library audio and ritual manifests need separate exports.
Session Pulse
Immediate, unblinded descriptions from guided sittings. They can reveal personal patterns. They cannot establish efficacy, neural entrainment or causation.
local only
A delayed follow-up is readyThis measures persistence separately from the immediate afterglow.
Personal Practice Map collecting
Each sitting contributes to several overlapping rows. Differences can reflect practice, guidance, narrator, sound bed, duration, intention, expectation, timing or ordinary variation. This map creates hypotheses. It does not rank settings or establish causation.
No check-ins yet. Enable Session Pulse in The Sitting when you want to measure a practice.
Sessions
Your listening timeline
Find a sound you selected, a playback request or a saved session record.
Events are not listening time. Completion markers can overlap detailed records.
What the timeline records
Selections, playback requests, completion markers and detailed sitting, journey or trial records are separate events. A selection or request does not prove you heard audio. Older activity has no verified stage. The activity log retains at most 300 events; detailed records have their own histories. Overlapping entries are not a total of completed sessions. Search covers retained readable records, including older matches, without changing them. History stays on this browser origin until you export it.
Activity history and recovery
Checking saved activity.
Drafts exist only in this page. Export before reloading. These downloads can reveal private activity and are not encrypted. They are not automatic restore packages or backups of your notes and audio.
Research tools
A classroom comparison
Join a class, prepare a protocol or review exported results. Opening a tool does not start audio.
Use the original code from your instructor, or one you create below. No names or emails.
not in a class
Join a classParticipant
Joining restores the sealed instrument settings and checks both sounds. Stop audio and turn ear calibration off first. Nothing plays automatically.
Once joined, use Trials for the assigned sessions, then return here to export your CSV. Leaving does not restore your previous instrument settings.
Set the console sound, stop playback and turn ear calibration off. Choose the plan, then create a code. Both sounds are rendered and checked before a code is issued.
Archive the generated code and share it with numbered slots assigned outside BINAURATUNE. Paste that same code above when reviewing results.
Review exported resultsInstructor
Paste the original class code above, then open participants' CSVs. Files are processed on this device. Invalid or incomplete records withhold the planned result.
No classroom files are open.
For learning study methods. Public codes do not prove identity, consent or treatment benefit. Participants choose whether to export their records.
Protocol rules & limitationsOptional reference
Plan before enrollment
Classroom Protocol v2 fixes one sound, one control, one primary rating outcome, a delivery route, duration, participant count and session schedule. Each four-session block contains two tested and two control sessions. Assign numbered slots without names here. The interface hides labels until ratings are saved; the public code and open source make this instructional masking, not cryptographic concealment from a technically motivated participant.
Digital checks, not identity proof
Code creation and joining each render both arms. The gate refuses more than 0.5 dB RMS difference per ear, nonfinite samples or a peak at the safety ceiling. Ear calibration must be off. A public code uses base64, not encryption or a signature. SHA-256 identities detect accidental changes and mixed protocols but do not authenticate the instructor.
Keep missing data visible
Participants complete the required playback check, run the balanced sessions and choose whether to export a minimized CSV. The pooler requires the exact v2 header, class identity, numbered positions, sealed settings, declared route and finite ratings from 1 to 10. It reconstructs assignments from the public seed. Invalid rows and duplicate positions are refused, not quietly included.
Interpret the result carefully
The planned result requires every numbered slot and complete four-session blocks without missing or refused rows. The exact blocked randomization test evaluates the sharp no-effect null, not the probability that a hypothesis is true. Neither a public code nor a browser can prove unique people, consent, independent recruitment, or protocol compliance outside the app. Personal ratings are not brain measurements.
Compare
What the leaders do better, and what nobody else does at all
Most comparison pages are advertisements with a table. This one begins with what the biggest products in this category genuinely do better than we do, because if we will not say it, you have no reason to believe the second half.
Where the big apps beat us
Recorded human content. Calm and Headspace employ studios, writers and beloved narrators; their libraries hold thousands of professionally produced hours. BINAURATUNE now offers the precise device narrator, one licensed human opening, or a measured numbered human cue pack that preserves a complete generated meditation and its timed silence. Preset-only VibeVoice and multi-engine Voicebox remain consent-gated local options. These production paths still do not replace a rights-cleared human library or its art direction. That remains a product requirement.
Native apps everywhere. They live on phones, watches and televisions with store presence and offline downloads. We are a page you can install (it works fully offline once installed), which is lighter and more private, and still not a store shelf.
Context adaptivity. Endel reads weather, heart rate, movement and location to reshape sound in real time. BINAURATUNE now makes a different trade. The Private Context Lens adds explicit room, movement and interruption facts, optional five-second local room and motion hints, and an optional standard Bluetooth heart-rate reading. Sensor hints never apply themselves, BPM remains observational, context expires on reload, and each recipe carries a complete decision receipt. The Doorway separately offers one transparent next action from a due follow-up, a plan you saved, a recently opened item, or the device hour, in that order. We still do not provide continuous weather or location adaptation. We do provide a concealed adaptive-versus-static test that can discover whether our additional context creates value rather than merely movement.
Sound aesthetics. Art-directed studios still beat synthesis for sheer beauty. BINAURATUNE now tests two complete masters through The Full Listen, including beauty, fatigue, repetition, pacing and ending quality, but a better quality gate does not create the instruments, performances and art direction that a professional source library still requires.
Where this category cannot follow
The most recent systematic review of the EEG literature found only five of the fourteen studies supported the entrainment hypothesis. Every product in this category, including this one, lives under that fact. The difference is what each does with it: the category's marketing calls its sound neuroscience-backed; we registered the review in our own evidence ledger, publish what each claim does not support, and hand you blinded trials so your own data outranks our copy.
A sham you cannot tell apart. Blinded conditions are acoustically matched within half a decibel per ear, measured, re-runnable from the footer of this page. Without a matched sham, a self-test measures expectation. Nothing else in the category ships one.
Nothing to cancel. No account, no subscription, no tracker, complete export and erasure, audited. The regulator currently re-examining this category's billing practices has no business with us, because there is no bill.
Active-control discriminability · sealed listening study
Can an exact real and active-control sound be identified before outcomes, expectations or perceived benefit can influence the answer? BINAURATUNE renders the selected pair through the production master chain, balances ten A-real and ten B-real trials, and conceals every answer until the fixed twenty-pair block is complete. Each clip must finish naturally once per pair. There is no trial-by-trial feedback and no optional stopping.
Pinned binaural reference
10 Hz binaural signal at a 220 Hz carrier versus its level-matched still control.
Production PCM will be rendered and measured before allocation.
Listening setup receiptChoose the physical output used for this block. Evidence transfers to a Trial only when the exact sound fingerprint and this declared route both match.
Choose the route you will actually use. No route is assumed.setup not declared
Headphones are required. A full block takes about six minutes.
Pair 1 of 20answers remain concealed
pure guesscertain
Primary endpoint: correct identifications out of twenty, tested against chance with a two-sided exact binomial test. A non-significant result does not prove the sounds are indistinguishable. A significant result shows that this listener could discriminate this exact pair under this playback setup. It does not identify the cue, measure brain activity, or validate any psychological outcome.
Named products above are named for what they do well; their strengths are real and we recommend them for what they are best at. The comparison each reader should trust most is the one this page cannot write: run your own blinded sessions in Trials and let the inference engine tell you what your data can and cannot support.
Care and Keeping
Listening and meditation care
Sound and meditation can be uncomfortable for some people. These precautions support informed choices; they do not establish that a session is safe or suitable for everyone.
Do not use the flicker
The Lab includes an optional visual strobe. Flashing light can trigger seizures in susceptible people. As a precaution, do not use flicker if you have epilepsy, any seizure history, or a first-degree relative who does. If you have a seizure history, ask your neurologist about sound-based sessions before trying them.
The flicker stays off until you switch it on, and it never starts on its own.
Check first if
If a health condition or past experience makes you unsure whether sound, meditation or visual effects are suitable for you, ask a clinician who knows your situation.
Do not use sessions while driving, operating machinery or doing anything that requires your full attention.
Unwanted responses
If you notice headache, increased ringing, anxiety, agitation or other discomfort during a session, stop. You can choose quiet or leave the session unfinished. Adjusting settings does not establish that continuing is safe.
Stop if listening worsens your tinnitus. Seek professional advice for persistent or worsening ringing or hearing difficulty. BINAURATUNE has not been shown to treat tinnitus.
Meditation is not risk free either
Some people experience anxiety or other unwanted effects during or after meditation. You can keep your eyes open, choose quiet or stop at any time. If practice becomes distressing, stop and return your attention to your surroundings. Seek appropriate support if distress continues.
Volume
Use a low, comfortable volume and take regular breaks. Hearing risk depends on sound level and duration. This app cannot measure the level reaching your ears, and its digital limiter does not guarantee safe listening.
Everything this knows about you
Counting.
Private, unencrypted JSON, including raw stored text. Library audio and ritual manifests need separate exports.
There is no account, no server and no analytics. Settings and records live in local storage; optional mastered WAVs and Continuum rituals live separately in the private IndexedDB library. Clearing site data destroys both. Download the JSON occasionally, download each offline master you cannot replace, and export ritual manifests when their order matters. The JSON restores ordinary records into another browser, WAV files remain the portable audio copies, and manifests preserve ritual order without containing audio.
Evidence
What the research actually says
Explore the findings, follow the sources and see the limits. A study result is not a promise of personal benefit.
The complete registry loads when you open Research.
Findings & sources
Evidence library
Search every claim, source and limitation. Review dates track maintenance, not scientific certainty.
A fresh review date means the source was checked recently under this policy. It does not upgrade the evidence grade, replicate the study, guarantee completeness, prove independence, or turn a finding into a personal outcome.
Review & verification tools
For reviewers and maintainers. Sign reviews, inspect releases and verify organization keys. Your work stays in this tab when you close the panel.
Tools load only when opened.
Context & further reading
What we decline to build
Eight-dimensional audio. Competitors sell this hard. We searched for a peer-reviewed literature under that name and there is none. What the term describes is slow rotational panning, which at the speeds those tracks use is ordinary perceptible movement rather than an illusion at all, and no study shows it helps relaxation or attention beyond plain novelty. That is not the same as saying it is useless: it is untested, which is a different thing from disproven. We would rather ship nothing than ship a claim we cannot support.
A personalised gamma frequency. This is the feature we most wanted to build. Everyone's gamma peak sits somewhere slightly different, modelling work suggests a two-hertz mismatch costs most of the phase locking, and we already run a validated gap-detection test whose threshold is known to correlate with individual gamma frequency. It looked like a clean line from a measurement we have to a number we could tune to. Two things stopped us. The correlation has only ever been established with EEG measured first and gap detection compared against it; nobody has published the inversion, so turning a millisecond threshold into a hertz value would be our extrapolation wearing the costume of a measurement. And the study most often cited for personalisation actually found that an auditory-derived individual frequency engaged fewer brain regions than plain 40 Hz, thirteen of sixty-eight against twenty-seven; the condition that won was task-derived and its own authors call it state-dependent rather than a stable trait. So there is no personalised gamma dial here. If we add one it will arrive as a Lab protocol wired into blinded trials, where your own data decides, rather than as a number we assert.
Closed-loop stimulation needs a measured loop. Published EEG-gated studies demonstrate stimulation that listens back: sounds are timed to measured neural phase. That differs from playing a fixed rhythm. There is no closed-loop toggle here because BINAURATUNE has no validated EEG phase-targeting pathway. Device buffering, transport delay, sound-output latency and artifacts must be measured before a future integration can claim precise timing. Overnight use also needs validated sleep-state detection in its intended setting. This remains hardware-dependent research, not a proven feature of this app.
A sleep protocol that promises better memory. This is the one we most regret leaving out, because the physiology is real and beautiful: clicks timed to the rising phase of a sleep slow oscillation reliably deepen that oscillation and the spindles riding on it, and that part replicates everywhere. The memory benefit is a different story. A meta-analysis of fourteen studies on overnight word-pair retention found that the year a study was published accounts for 91.8 percent of the variation between them, which is a polite way of saying the effect has been shrinking every year since the first report. Several careful double-blind replications now enhance the brain rhythm exactly as advertised and find no behavioural memory gain at all. A shrinking effect with a perfect time trend is what publication bias looks like from the inside, so we are not shipping an overnight memory protocol, and we would treat anyone who does with suspicion.
A creativity preset. Alpha enhancement really does appear to help divergent thinking, and it is one of the better-replicated cognitive findings in the neighbourhood: 10 Hz stimulation improved it where 40 Hz did not, and parieto-occipital replications have since followed. Every one of those studies pushed alpha with electrical current through the scalp. We push it with sound, and nobody has shown that sound does the same thing. Alpha is here in several states and you are welcome to use it while working on something open-ended, but we will not put the word creativity on a card and let you assume the evidence transferred, because the transfer is precisely the untested step.
Infrasound layers. Sub-audible content below about 20 Hz is a popular addition to calming tracks. The one recent controlled test we could find, thirty-six people in a randomised design, found that infrasound nobody consciously detected still raised salivary cortisol and worsened mood. That is the opposite of what a calm preset is for, so there is none here.
Where a simpler answer won
Paced breathing here runs at a fixed six breaths a minute, one cycle every ten seconds, with no calibration step first. Some apps ask you to find a personal breathing rate before you begin. We kept it simple.
So the breath tide asks you nothing beforehand. You can start breathing with it straight away.
Claim
Standing
Reading
The brain's electrical activity measurably follows rhythmic sound (the auditory steady-state response)
A single session of 40 Hz stimulation improves attention or memory in healthy people
Unsupported
A 2025 study entrained gamma in real time, confirmed it on EEG, and then found no gain on perception, attention, short-term or long-term memory. Worth holding onto whenever you read a gamma claim: driving the rhythm is not the same as improving the thinking, and BINAURATUNE can generate an acoustic rhythm, but it does not measure your EEG or establish that your brain follows it.
Slow-paced (resonance) breathing raises vagal tone and calms the stress response
Supported
HRV-biofeedback literature; our Breath Tide couples it to sound. That pairing is our hypothesis, not settled science.
Cymatic "healing": that sound-made patterns cure illness in body tissue
Unproven
No controlled evidence. The physics above is real; the leap to therapy is not yet earned. Vibroacoustic therapy is the nearest credible relative.
Hemi-Sync produces whole-brain synchronization and trains out-of-body states
Unproven
Monroe's signal layering is a real design idea and the CIA's 1983 Gateway analysis took it seriously, but rigorous independent EEG evidence never materialized. The Ladder in our Lab is a testable rebuild.
Binaural beats entrain cortical EEG rhythms to the beat frequency
Contested
Systematic review, 2023: of 14 EEG studies, 5 were in line with entrainment, 8 contradicted it, 1 was mixed. The brainstem response is real; the cortical claim is open.
Polysomnography crossover, SLEEP 2024: a 0 to 3 Hz sweep over a 100 Hz carrier cut sleep latency versus sham (n=10). Our Undertow journey is this design.
Stimulus craft is the cheapest efficacy lever: carrier, masking, and onset choices change outcomes
Supported
Factorial study, Sci. Reports 2025: gamma beats over a 340 Hz carrier with white noise beat other combinations for attention. Our Gamma 40 preset and comfort engine apply this.
Binaural beats over pink noise shorten EEG-verified sleep onset
Early
Moscow crossover, n=14: sleep onset 429 vs 809 seconds in responders; monaural beats failed the same test. Our Hypnagogia preset is this recipe.
Sound clicks phase-locked to deep-sleep waves deepen slow oscillations
Five-element (wuxing) music heals the organs each tone is mapped to
Unproven
The tone-organ map is a 2,000-year-old correspondence system, not acoustics. Trials show mood and quality-of-life gains over silence but rarely over generic music: meta-analysis of 22 cancer studies · Liao et al., 2013. Our Temple atmosphere is the testable version.
Monroe's Spatial Angle Modulation: a tone moving through a stereo arc entrains where beats cannot
Unproven
The engineering is real and public in an abandoned 2011 patent, and decoupling rate from detuning genuinely permits gamma-rate spatial rhythm. But the only published evidence is the Institute's own, and its quantum-mind rationale is speculative. Our SAM mode rebuilds the method so it can finally be tested blind.
EEG turned into music heals (Russian bioacoustic correction, Music of the Brain)
Unproven
A real St Petersburg and Moscow engineering lineage with inventor-run trials claiming 80 to 93 percent response: Levin, 1998 · Fedotchev, 2018. Interesting idea, unproven claims, and impossible without EEG.
Expert meditators self-generate extraordinary gamma synchrony (the monk finding)
Supported
Lutz et al., PNAS 2004: practitioners with 10,000+ hours produced sustained 25 to 42 Hz synchrony at will. The state is real and trained, not sound-induced; our gamma modes offer scaffolding, not a shortcut.
Savant-like skills can be unmasked in ordinary people by suppressing the left temporal lobe
Early
Snyder et al., 2006 (numerosity, rTMS) and Chi & Snyder, 2012 (tDCS, insight). Magnetic and electrical, never sound; small samples, mixed replication. The Numerosity Window is our probe for any audio analog.
Clairvoyance and telepathy demonstrated by ganzfeld and remote-viewing experiments
Consciousness is quantum (Orch-OR), so rhythmic sound taps quantum coherence
Speculative
Penrose and Hameroff's microtubule theory has intriguing new threads (anesthetics acting on microtubules, 2025 supportive review) and severe critics (decoherence in femtoseconds). No link from audible sound to quantum brain states exists. Randomized draws use browser cryptographic randomness. You can optionally add ANU-supplied bytes in Trials, but BINAURATUNE cannot independently verify their physical origin. Randomness provenance does not establish a consciousness effect.
Specific frequencies reliably induce sleep, meditation, or altered states in everyone
Unproven
Anecdote-rich, trial-poor. In the same lab some listeners respond strongly and others not at all; individual variation is the rule, which is why this instrument keeps handing you measurement tools.
Every row below is measured from rendered PCM samples by independent analysis (zero-crossing frequency estimation, Hann-windowed spectral estimation, envelope extraction), never from the oscillator settings that generate the signal. Green passes, amber is an informational warning, red fails. Layer rows exercise generators in isolation; output rows exercise the true master path your ears receive.
Application script beginsmeasuringNavigation start to application execution, including loading and parsing.
Setup callsmeasuringSum of timed setup calls. Gaps between calls are excluded.
Ready to usemeasuringNavigation start until setup completes and controls unlock. This is not a paint measurement.
First contentful paintnot reportedActual browser paint entry for the first text or image. It does not mean all tools are ready.
Setup wall timemeasuringFirst setup call through completion, including scheduled gaps.
Longest setup callmeasuringFind the next task to split using measured call duration.
Document elementsmeasuringActual elements in this loaded build.
Document bytesbrowser withheldDecoded navigation bytes when the browser exposes them.
Section containmentmeasuringRendering optimization support and active section count.
Startup timings are direct browser observations, not pass claims. Device speed, cache state, extensions and developer tools can change them. Section containment reduces work below the viewport without removing content from search, keyboard focus or accessibility APIs.
Optional tools load when opened. No audio starts automatically.
Listen comfortably.Keep the volume low and stop if uncomfortable. Do not use while driving. If you have a seizure disorder, avoid flashing light and consult your clinician before rhythmic stimulation. Not medical care. Read the listening guide.
Updated BINAURATUNE tools are ready. Save any unsaved work, then reload when this session is finished. Nothing will refresh itself while you are listening.Reload now
let the field fill your vision · imagery arrives in minutes · tap to end
The Index
Find your next practice
Search the complete local BINAURATUNE catalog by title, goal, method, duration, listening route, or scientific maturity. Opening a result never starts sound or speech.
Building the local index
Search relevance is not a claim that a practice works. Favorites and recents stay on this device and are included in your BINAURATUNE data export.
Sixty seconds to your first session
1. Put on headphones; the default method needs one tone in each ear.
2. Press play and let the 10 Hz alpha rhythm run. It sounds like a gentle wobble; that's the point.
3. After a few minutes, notice: shoulders, jaw, breath. Then explore the States below, or test yourself in Trials.
Install and offline
Carry BINAURATUNE
Install this web app for a dedicated window, a home-screen icon and a locally cached core. This is an installed web app, not an app-store-native release.
Opening asChecking this window
Offline coreChecking app files
Available with the app files
The complete sound instrument, journeys, experiments and deterministic browser composer.
Meditation manuscripts, timers, local history, trials, reports and saved settings.
No account, tracking connection or remote data store is required.
What still depends on your device
Spoken browser narration depends on voices installed by your operating system and may not be offline in every browser.
AI voice and music generation need the configured local engine or online provider. Saving the app does not install AI models.
Your recordings and generated tracks are stored separately. Keep exported backups; browser storage can be cleared.
Bluetooth, spatial processing and accessibility settings can still alter playback after BINAURATUNE generates its stereo signal.
Native Android receiver preview
A compiled Android debug preview now supports two strict paths. It can recreate seven protocol methods as seek-stable stereo PCM, or validate and preserve one complete BINAURATUNE mastered WAV with narration, music and cue timing. Both use Android background media controls, preserve position, and pause headphone-only audio after route loss.
Supported today in the compiled preview: Binaural, Monaural, Isochronic, SAM, Blend, Purist Comb and Nested with a sine carrier.
A Session Mastering Room WAV is independently checked from its actual samples and embedded record, copied to private app storage, then prepared without autoplay.
Unsupported stages are refused before playback instead of being silently removed.
The debug APK is reproducibly built, package-inspected and debug-signed. It is not release-signed or published. Measured PCM parity, battery, interruption and physical-device testing remain required.
Check for the preview matching this app build before downloading. Nothing starts automatically.
The check reads a small build receipt and requests file headers from this website. It does not download or install the APK, verify its signature, or test your phone. Preview files need a connection and may not be hosted with every copy of BINAURATUNE. This is a debug preview, not an app-store release.
How to move your session to Android
Choose the file that matches what you want to hear. Neither path is cloud sync, and opening a file never starts playback automatically.
Protocol file: tones only
Select Phone app below. Review every listed compatibility item. Adjust the instrument deliberately if you want to use this path; BINAURATUNE does not remove effects for you.
When the destination contract is ready, choose Export session file. This contains settings and session position, not a recording of narration or music. An active source session pauses only after the export request succeeds.
Transfer that file yourself, then open or share it with BINAURATUNE Receiver on Android. Inspect the received settings and press Apply, then Play when ready.
Mastered WAV: narration and music
In The Sitting, open the Session Mastering Room. Prepare the complete narration and music, review your permissions and listening checks, then choose Export current session WAV.
Use Send last master to Android when file sharing is supported, or transfer the unchanged exported WAV yourself. This preserves the rendered mix instead of asking the phone to recreate each effect.
Open it with BINAURATUNE Receiver. The receiver checks the audio and embedded production record, then prepares it without autoplay. A generic WAV, MP3 or editable session backup is not a Session Mastering Room WAV.
First obtain the matching debug preview above. Installation is your choice and requires a compatible Android device. Headphone-only sessions still require isolated stereo headphones. The phone must validate each file; a successful web export is not phone acceptance. Active guided sittings and blinded trials cannot be moved as a live protocol.
Move one session between devices
Export the exact requested protocol, elapsed position and remaining time as one integrity-checked file. Another browser can play the full schema today. The Android debug preview accepts only the subset it can reproduce exactly enough to test, and the destination check below will tell you when it would refuse the current setup.
Destination
Fallback length for an untimed instrument
Destination contractChecking this protocol
Incoming handoffNo file opened
Open this panel after setting the instrument or starting a Journey. No audio or personal history is included.
The file transfer is manual and local. It is not cloud sync, wireless handoff or an app-store release.
A SHA-256 digest over RFC 8785 canonical JSON detects changed content when the digest has not also been recomputed. It does not prove who created a file and is not protection from a determined attacker. Imported files never start sound without a fresh press. Trials, active guided sittings and photic state are never transferred.
Checking the installation path available in this browser.
Set up your comparison
Review each section. Nothing starts until you choose Begin.
Protocol: ·
Pre-exposure acoustic identity
1 Session
Length:
Control:
Removed control stills the rhythm. Detuned control changes its rate. Audible differences can reveal either condition.
Starting consumes a concealed position, even if you later stop without an outcome.
How assignment works
Personal trials use blocks containing two tested and two control sessions in hidden random order. If the open-label silence arm is enabled, the block contains two of each condition across six positions. Starting a condition consumes its hidden position even if the session is later abandoned.
2 Playback
Choose the output you will use. Your choice is a declaration, not automatic device detection.
Choose the route you will actually use. No route is assumed.Playback receipt details
A browser cannot reliably identify whether those samples reach headphones or speakers. The declared route and latest listening check are saved with the consumed position.
delivery not declared
3 Your plan
Choose an outcome and stopping rule. The first allocation seals these choices for this exact protocol.
Planned completed sessions
Smallest worthwhile real minus sham change
Your worthwhile difference on the 1 to 10 rating scale, not a validated clinical cutoff.
Minimum separation between exposed sessions
Research does not establish a universal washout time for BINAURATUNE audio. Choose an interval for your question; the default is 24 hours.
The plan will be sealed before condition allocation. Analysis is two-sided even when the expected direction is positive.Plan identity & precision forecastnot sealed yet
Blinded precision forecastReference scenarios will appear here.
This planning ruler never reads an interim condition contrast.
4 Before you listen
How do you feel right now?
Your expectations
Answer all six questions honestly, including low expectations. Ratings cannot establish cause by themselves.
About these questions
These are the Credibility and Expectancy Questionnaire items used by BINAURATUNE. They record what you think about a session and how much change you expect or feel will occur. Recording expectations does not remove expectancy effects or prove that sound caused a change.
Optional paired tests
Use the same device and input method, and keep the tab visible. These measure task performance, not intelligence or brain entrainment.
Session in progress
Settle in. The sound will fade out by itself, then we'll ask how it went.
10:00
If you end early, the assignment record is kept and marked abandoned. No completed outcome is added.
How did it go?
First, record what you think you heard. This answer locks before the outcome ratings appear, and the condition remains concealed until everything is saved.
Which condition do you think it was?
pure guesscertain
This is a post-exposure concealment check, not proof of blinding. A guess can reflect an audible cue, a preferred answer, or how effective the session felt. BINAURATUNE reports response patterns and uncertainty rather than declaring a blind successful from one cutoff.
pair 1 · arm A
How did this guidance structure fit?
The condition remains concealed until both arms in this pair are rated. Rate the session you just experienced, not the idea behind it.
definitely notdefinitely yes
not at allvery much
too little roomample room
not at allrepeatedly
not at allvery tiring
If practice is increasing panic, agitation, traumatic memories, detachment, pain, or loss of functioning, stop. A preference experiment is not a reason to push through a difficult reaction.
Move every rating once before saving. The midpoint is valid, but touching it confirms that it is your answer.
before the sit
A quick private check-in
Place each marker where you are right now. These are momentary descriptions, not diagnoses or a test you can pass.
tense or uncomfortableat ease
difficult or unpleasantpositive
scatteredsteady
very uncomfortablevery comfortable
would not repeatwould repeat
Open your eyes, feel the support beneath you and let the room back in. Do not push through a practice that leaves you distressed or detached. If the reaction is severe, persists, or affects daily functioning, stop using the practice and seek qualified support.
not helpfulvery helpful
would not choosewould choose
Pause this practice if a difficult reaction is continuing. Orient to the room, feel the support beneath you and contact someone you trust. If the reaction is severe, persists, or affects daily functioning, seek qualified support.
Single ratings contain substantial measurement error. BINAURATUNE waits for repeated complete pairs before showing method patterns.
Check-in saves are confirmed before this editor closes.
Keep this page open if saving fails. Use the save button to retry these ratings, or export the draft before deliberately discarding it. A download may contain private health-related reflections.
The reveal
·
Reaction test
When the circle ignites, click it (or press space) as fast as you can. Seven rounds; too-soon clicks repeat the round.
click anywhere in the box to start
The Chordwright
Compose your own layered protocol: up to three binaural dyads sounding at once, the same machinery The Ladder climbs on. Sound it live, then take it straight into a blinded trial; sham strips every beat while keeping every tone. As Hymnal preserves the ratios between your voice rates, and the main beat dial scales them together.
The Numerosity Window
Dots flash for six tenths of a second; estimate how many you saw. Six rounds. Savants do this without counting, and magnetic suppression of the left temporal lobe briefly improved it in ordinary people. If any sound state moves your score, this window will catch it; expect noise before you trust any trend.
press begin
The Oracle Trial
Twenty draws. Before each guess, the instrument has already selected one of four symbols using browser cryptographic randomness. Chance averages five hits, not five on every run. A high score in a short run does not establish clairvoyance; repeated, prospectively planned comparisons are needed to distinguish a pattern from chance and selective reporting.
press begin
randomness: browser cryptographic randomness
The Six Breaths
Six syllables, eighteen slow breaths, about four minutes. Breathe in through the nose for four counts; then exhale long and soft for eight, letting the syllable shape the breath. The organ names are the traditional map, shown as tradition, not physiology.
ready
press begin
Resonant Breath
Follow one ten-second cycle: breathe in gently for four seconds, then breathe out for six. Do not take the largest breath you can. Quiet and comfortable is the target.
Choose a length, then begin.
Stop if you feel lightheaded, air-hungry, numb or uncomfortable, and return to ordinary breathing. This runner does not ask you to hold your breath.
The Strange Clock
Tap with two streams of apparently random clicks. One stream contains a hidden short-range timing structure. The other uses the exact same intervals in shuffled order, preserving their mean, spread and histogram. The order is concealed until both blocks are complete.
This is a simplified replication and extension of a 2025 online tapping study. It measures whether your motor timing exploits statistical structure. It does not measure intelligence, altered consciousness or a latent ability.
Two blinded blocks, about ninety seconds total. Keep this tab visible.
The Rhythm Gate
Compare three precisely timed drum streams at 1.5, 4, and 9.5 pulses per second. Their numeric labels and order stay concealed until all three blocks are complete. The rhythms themselves remain perceptibly different, so this is condition-label blinding, not an indistinguishable sham.
A 2026 Scientific Reports study used these exact rates for ten minutes each in forty listeners. At group level, 9.5 pulses per second produced a higher altered-experience score than 1.5, while 4 was not uniquely strongest. Stronger EEG tracking at 4 Hz predicted stronger altered experience in some people. This browser version records experience and time estimation but has no EEG, so it cannot measure neural tracking or certify an altered state.
Three concealed conditions. Use a comfortable level, close your eyes if that feels safe, and keep clocks out of sight.
Different from ordinary waking experience
Absorption in the sound
Physical comfort
The Noise Ladder
First, a short adaptive staircase finds a complex tone near your own hearing threshold on this device. Then a concealed six-level comparison asks which of three intervals contains that tone. The conditions range from quiet through minimal white noise to a masker as strong as the signal.
Low-level noise can strengthen a neural response to a weak sound, but behavioral benefits are inconsistent and highly individual. This test measures detection accuracy only. It does not measure EEG, brainwave entrainment, focus, or a change in consciousness. Browser audio cannot calibrate sound pressure level, so the result is valid for this device, output level, and listening setup only.
Use headphones somewhere quiet. Set a comfortable device volume and leave it unchanged for the entire comparison.
The Flux Duel
Two concealed versions of the same harmonic texture move at exactly 2 Hz. One moves brightness while holding energy nearly steady. The other keeps brightness steady and pulses amplitude. Tap with the pulse, then rate clarity, comfort, and how well the sound could sit behind another task.
A human EEG study found that spectral flux in natural music predicted neural synchronization better than the amplitude envelope. This synthetic comparison is a new translation of that result, not a replication. It measures tapping consistency and preference. It does not measure EEG, consciousness, focus, or treatment benefit. Labels remain A and B until both blocks are complete, although the two mechanisms may still sound different.
Use a comfortable fixed device volume. Speakers or headphones both work.
Pulse clarity
Physical comfort
Fits behind another task
Repeated-session evidence plan
Seal the attempt count, spacing, and smallest tap-lock difference worth following before the next concealed allocation. At most two attempts can enter the plan on one UTC date. Started attempts consume slots, interrupted attempts remain missing, and no effect estimate appears before the stopping boundary.
A personal plan needs no study file. A cohort contribution must be bound to a public charter and externally assigned slot before the first attempt.
No repeated-session plan is active. Ordinary duels remain descriptive practice.
One silence, two prediction structures
The Omission Gate
Two concealed streams contain the same tones, the same event times, and the same missing events. In one stream, each position in the four-tone cycle owns a stable pitch. In the other, those exact pitches are shuffled within each cycle. Press the pad whenever a tone should have arrived but did not.
Human EEG studies find larger omission responses when sound identity is predictable. Recent animal recordings also show omission responses scaling with tone probability and sequence regularity. This browser comparison asks a narrower behavioral question: does stable pitch identity improve your detection of the same silences? It measures button presses, not EEG, predictive-coding neurons, consciousness, attention benefit, or treatment.
Primary measurehit rate minus false alarms
Held fixedPCM level, tones, timing, omissions
Changedpitch identity predictability
Two concealed blocks, about two minutes total. Speakers or headphones work. Keep this tab visible.
Personal evidence plan
Seal the number of attempts, spacing, and smallest difference worth detecting before the next concealed allocation. Every started attempt consumes one slot. Interrupted attempts stay visible, and BINAURATUNE withholds the planned estimate until the target is reached.
No personal evidence plan is active. Runs remain useful practice, but they are not part of a sealed repeated comparison.
Same composition, three pitch references
The Tuning Trial
Hear one deterministic composition three times in a concealed order. Every note, chord, rhythm, envelope, stereo position, duration, and measured RMS stays fixed. Only the global pitch scale changes.
The conditions are A4 = 440 Hz, A4 = 432 Hz, and C5 = 528 Hz. The last condition is equivalent to A4 = 443.99 Hz. It does not add a special 528 Hz tone. Published tuning studies are small, use different music and outcomes, and do not establish healing, DNA repair, neuroplasticity, or a universal best tuning. This 45-second-per-arm run measures your immediate preference and state only.
Primary ratingpleasantness
Held fixedcomposition and RMS
Changedglobal pitch scale only
Use one comfortable fixed volume. Speakers or headphones both work.
Move every rating before continuing. Brightness is a manipulation check, not part of the preference score.
One audio buffer, two breathing phases
The Body Clock
Listen to the exact same two-minute pitched pattern twice while following a quiet visual breath guide. In one concealed arm, missing downbeats fall during paced exhalation. In the other, the identical audio timeline is replayed while the guide places those same moments during paced inhalation.
Press the large pad or Space whenever you detect a missing downbeat. The primary result is omission-detection sensitivity, hit rate minus false-alarm rate. The browser does not measure your breathing, so the comparison is meaningful only if you follow the guide comfortably. This is a behavioral pilot, not EEG, respiratory measurement, treatment, or proof that one breath phase changes consciousness.
Held identicalevery audio sample
Changedvisual breath phase
Primary outcomedetection sensitivity
Two concealed arms, about five minutes including breathing lead-ins and ratings. Speakers or headphones both work.
Move every rating before continuing. Low guide-following scores remain in the record and are flagged rather than silently excluded.
The Sweet Spot
Five coupling levels, thirty seconds each, in random order. After each one, rate how locked-in the rhythm felt, meaning how much it pulled you rather than simply how loud it was. Theory predicts your curve peaks somewhere in the middle. If it climbs all the way to the top instead, that is a real result too, and it is yours.
press begin
The Carrier Fit
Four identity-concealed segments compare your saved carrier with a deterministic alternative while the 40 Hz pulse, waveform and production path stay fixed. Each carrier appears twice in a balanced ABBA or BAAB block. BINAURATUNE measures rendered PCM first and attenuates only the louder arm until the per-ear level difference is within 0.25 dB. Rate comfort as the primary outcome and desire to continue as a secondary outcome.
The carrier pitch is audible, so this is identity concealed rather than fully blinded. One block is descriptive. A repeatable preference needs at least three different UTC dates and may still resolve to no reliable separation. This tests listening fit, not EEG, entrainment, cognition or treatment benefit.
No comparison registered yet.
press begin
Recovery is local to this page. Unsaved exports can include existing fitting history needed to review the change. Close other BINAURATUNE tabs before retrying. Downloads do not merge or repair records automatically.
Prospective personalization test
The Route Fit Charter
Does your measured binaural masking release predict whether a binaural or monaural 10 Hz beat feels more comfortable? BINAURATUNE freezes the hearing result before exposure, holds the 300 Hz carrier and target rate fixed, conceals route identity, and lets either route win.
BMLD is an established binaural hearing measure. It has not been validated as a predictor of binaural-beat comfort, audibility, entrainment, benefit, or treatment response. Monaural beating is physically present in each ear and will often sound more obvious, so this is identity concealed rather than fully blinded.
Measure binaural release in The Phase Key on two different UTC dates before registering this comparison.
Frozen BMLDnot registered
Predictionnot registered
Complete blocks0
Different dates0 of 3
No Route Fit Charter is active.
Recovery is local to this page. Unsaved exports can include existing fitting history needed to review the change. Close other BINAURATUNE tabs before retrying. Downloads do not merge or repair records automatically.
The Metronome Inside
An objective test with no ratings in it. Tap your natural steady beat fourteen times. Listen for one minute to a blinded condition: a 2 Hz pulse, or its loudness-matched sham, chosen by hardware coin. Tap again. If rhythm entrains your motor system, your tempo drifts toward 2 Hz after the real condition and not after the sham, and the only data is your own tap timestamps. The tapping literature says motor synchronization in this range is robust; whether sixty seconds of OUR pulse leaves a measurable trace in YOU is exactly what one run cannot and many runs can answer.
press begin
Three measurements, no hidden score
The Structure Profile
Test three different parts of rhythmic behavior: whether you use serial timing structure, whether you infer a pulse that is absent from the sound, and how variable your unprompted tapping is.
BINAURATUNE reports the measurements separately. It never combines them into a talent, consciousness, intelligence, diagnosis, or latent-ability score. Relationships across tools remain exploratory until they repeat across at least eight shared UTC dates.
Serial timingnot measured
Absent beatnot measured
Free-tap variabilitynot measured
Complete each measurement on repeated days to test the planned relationships.
How the absent-beat probe works
Both concealed streams contain the same number of identical sounds, the same duration, the same level, and zero event-train energy at the 2 Hz target. One stream repeats a syncopated pattern that can imply the missing pulse. The control redraws an equal-energy pattern every four seconds. Tap the steady beat you perceive, if any. The result compares phase locking to 2 Hz, not brain activity.
The absent-beat probe has two concealed thirty-two-second streams.
Your Own Tempo
Tap the pad at whatever pace feels most natural, the speed you would walk or nod at when nobody is watching. No sound, no target, thirty seconds. Then the instrument will tell you your spontaneous tempo and offer rhythms built from it.
tap here, or press the space bar
press begin, then tap freely
The Vigil
A three-minute brief Psychomotor Vigilance Test. Wait for the field to light, then press it as quickly as you can. The delay changes randomly from one to four seconds. Pressing early counts as a false start.
This follows the validated PVT-B timing and its 355 ms lapse threshold. It measures behavioral alertness, especially sensitivity to sleep loss. It does not measure intelligence or prove that a sound changed cognition. Compare runs only on the same device and browser, because screens and input hardware add latency.
Three uninterrupted minutes. Keep this tab visible.
The Smallest Silence
You will hear three bursts of noise in a row. One of them has a tiny silence cut into its middle. Say which. It gets harder each time you get it right, and easier when you miss, until it settles on the shortest gap you can reliably hear. Takes about three minutes.
The Undecided
A repeating low-high-low triplet can settle into three readings: one galloping sequence, a high line in the foreground, or a low line in the foreground. Report what you hear at the start, then press again only when it changes. Two minutes.
This is a descriptive map of one listening period. It does not measure consciousness, attention quality, personality, hearing health, or a stable trait. Speakers or headphones both work.
The Stream Atlas
Compare the same low-high-low triplet at three pitch separations. Each one-minute arm is labeled A, B, or C until every arm is complete. Report one galloping sequence, a high foreground line, or a low foreground line at the start, then press again only when your reading changes.
The order is randomized and the waveforms have identical sample RMS, timing, duration, and output processing. Pitch separation is audible, so this is not truly blinded. It tests your reports in this session, not consciousness, attention, hearing health, personality, diagnosis, or a stable trait.
Press begin when ready. Speakers or headphones both work.
The Level Turn
Four short arms test whether an upward level change reorganizes an ambiguous low-high-low sequence more than a downward change. Each arm is called A, B, C, or D until the full comparison ends. Report one galloping sequence, a high foreground line, or a low foreground line at the start, then press again only when your reading changes.
The sequence identity is concealed, not blinded, because level changes are audible. Use one comfortable device volume and do not adjust it during the comparison. The signal uses a 12 dB digital step, not a calibrated sound-pressure level. This is a behavioral translation of a published laboratory effect, not a hearing, neurological, personality, or consciousness test.
Press begin when ready. Speakers or headphones both work.
The Second Look
Register eight future dates before asking whether the Level Turn direction recurs. The app selects the first eligible complete run on each UTC date, keeps directions concealed, and stops at eight.
not registered
This prospective plan uses one fixed outcome and an exact directional sign test. It does not validate a trait or personal neural effect.
The Release
Tap with a two-pulse-per-second tone, then keep tapping when the sound stops. Four concealed endings differ only in the acoustic phase immediately before the same short safety fade.
This tests post-sound tapping behavior. It does not measure EEG, neural persistence, consciousness, meditation depth, intelligence, talent, diagnosis, treatment response, or a stable trait. Keep one comfortable volume and use the same finger for every arm.
Press begin when ready. Speakers or headphones both work.
The Missing Thread
A steady tone is briefly interrupted. Sometimes noise covers the interruption so well that hearing restores the missing sound. Decide whether the tone seemed continuous or interrupted. Twenty-four concealed trials take about four minutes.
This measures an auditory continuity illusion, not hearing loss, intelligence, suggestibility, or consciousness. Speakers work, but use the same quiet setup each time.
Press begin when the room is quiet.
The Half-Octave
Each trial plays two octave-layered tones separated by exactly half an octave. The physical relation is deliberately ambiguous. Report whether the second tone seemed higher or lower. Twelve starting pitch classes appear twice in a concealed order.
There is no correct answer. The result is your direction profile in this room, on this device, at this moment. It does not identify dialect, musical talent, personality, or neurological status.
Press begin when ready. Headphones and speakers both work.
The Coupled Voice
Wear headphones and softly whisper ta with each synthetic syllable. The pace rises from 4.3 to 4.7 syllables per second. BINAURATUNE measures the microphone recording after each run, then immediately discards the audio.
This is a browser pilot of the explicit accelerated speech-synchronization task, not the published stimulus. It reports direct, half-rate and third-rate coupling separately. It does not measure brain activity, language ability, intelligence or pathology.
Before starting Use wired or low-latency headphones in a quiet room. Keep playback comfortable. Do not turn it up to hide your own voice. Whisper rather than speaking aloud. The first thirteen-second check measures room noise, listens for headphone leakage while you stay silent, then confirms that whispered onsets are detectable. Microphone permission is used only during that check or an active run.
Calibrate this microphone, then complete two measured runs of about 80 seconds each.
The Afterimage
Ten seconds of noise with a spectral hole in it, then silence. Most people hear a faint pure tone in the silence, pitched inside the hole, made entirely by the auditory system. It is the safest way to briefly experience what a phantom sound is.
The Phase Key
Three noise bursts, one hiding a faint tone. Two rounds of the same task; in the second, the tone is secretly phase-inverted in one ear. The difference between your two thresholds is the binaural masking level difference, a real measure of the ear-timing machinery binaural beats depend on.
Playback check
The audio this instrument generates is verified at the sample level. Everything after that is outside the browser's control: Bluetooth codecs, mono accessibility settings, spatial audio, and speakers can alter or collapse the channels. This check repeats two listening tasks, preserves the stereo stimulus through the output graph, and binds the result to one declared headphone route for thirty minutes. Where the browser supports it, any observed change to the available media-device environment invalidates the receipt and stops active research sessions. This does not identify the active output device and cannot observe every output or processing change. No hardware identifier or device inventory is stored. It is an operational listening check, not a hearing test or hardware measurement. Headphones on, comfortable volume.
press begin
Coherence probe
Headphones required. Each round plays three moments of sound; in every moment both ears hear a tone at once, sometimes high, sometimes low. Count every high tone across both ears, then answer. Five rounds. Dichotic tasks like this are historically linked to interhemispheric transfer, but be honest with yourself about what a score means: with five rounds the score moves in 20-point steps, and a one-step swing is attention and practice, not rewiring. Run it before and after The Ladder, many times, over days.
press start when ready
The Carrier Map
Seven carriers, fifteen seconds each, all carrying the same 10 Hz binaural beat. After each, rate how vividly you perceived the beat. Headphones on; closed eyes help. 528 Hz is in the set on purpose, so the solfeggio legend gets its fair blind hearing.
press begin
Compose a journey
Your draft stays in this tab when you close. Saving or confirming Start over clears it. Reloading the page loses an unsaved draft.
Discard this draft? Saved Journeys and recordings remain.
Personal sound direction
Soundscape Fitting Room
Compare four production atmospheres for one real use. Every candidate comes from the same engine, level dial and output ceiling. Source names stay hidden until you finish, so a label such as rain or temple cannot cast the vote for you.
Natural sound can support positive affect and stress recovery, but preference also depends on the listener and context. This is a personal aesthetic fitting, not a treatment test or a formal ISO soundscape study.
What is this soundscape for
A restorative comparison that includes water, shelter and low-event textures.
not startedlisten for at least 8 seconds
Choose a goal, then build the fitting. Keep one comfortable volume for every scene.
Rate only what you heard
Eventfulness is reported separately and does not raise or lower the fit score. A lively scene is not automatically better or worse.
Sources revealed
What the fitting controls
It chooses only the atmosphere type. It does not change your beat, carrier, method, volume, pulse shape, reverb or trial condition. The production engine already uses per-scene loudness trims, but a home fitting is not a calibrated laboratory comparison.
No soundscape fittings stored on this device.
Concealed music personalization
The Long Listen
Compare your current Hymnal direction with the untouched palette original. Both arms keep the same composition seed, root, target beat, carrier-pair spacing, master chain and volume. Only motion, brightness, density and stereo width may differ.
This is a personal preference and endurance check. It does not measure neural entrainment, meditation depth, focus, sleep, healing or consciousness.
Time per arm
Held fixedtarget and composition
Primary judgmentdesire to return
Stored hereratings and recipes, never audio
not startedcomplete the selected time
00:00 / 10:00ready
Set the Director first, then build the comparison. Keep one comfortable volume for both arms.
Rate only the arm you completed
Desire to return is the primary judgment. The comfort score also includes beauty, comfort, low fatigue, low repetition and low distraction. Pulse audibility is reported separately and never treated as proof of a brain effect.
Directions revealed
What this comparison can establish
A result can tell you which of these two directions you preferred on this device, at this time, for this listening duration. One run cannot establish a stable trait. Repeat on different days before treating a winner as your durable profile. The one-minute option is a preview and is not an endurance test.
Why the test is concealed
Palette labels and the word customized can change expectations. Randomized Arm A and Arm B labels reduce that cue. The directions remain audibly different because those differences are the object of the test, so this is concealed presentation rather than guaranteed blinding.
