MORTIS. Frequencies

Mortis.

Frequency science meditation. 30 seconds to your first protocol. No sign-up required.

Already have an account? Sign in →
What this is for
01 · Stress reset

Say you're overwhelmed. The AI picks alpha (8-12Hz) for parasympathetic activation. Your nervous system has a name for what you're feeling.

02 · Focus protocol

Say you need to get into work mode. The AI picks 40Hz gamma for neural binding speed. Used pre-work, pre-competition, or before anything requiring sharp thinking.

03 · Sleep prep

Say you can't wind down. The AI picks delta (1-4Hz). Run it through a sleep track and your nervous system has a documented pathway to sleep onset.

04 · First time

No prior meditation experience needed. Say anything honest about your current state. The AI reads tone, pace, and content to infer your nervous system condition. No voice data is stored.

05 · Skeptic test

Take an HRV reading before. Do the session. Take another one after. The delta is your data point. One session is enough to decide if the mechanism is real for you personally. We do not ask you to trust anything you haven't measured.

Better results with a better intake

Say what you're actually feeling. Vague gets generic.

The more specific your description of your current state, the more specific the protocol recommendation. These examples show the difference.

Generic
"I want to meditate."

Claude gets 10Hz alpha or the Schumann resonance by default. It's not wrong, but it's not calibrated to anything specific about your current state.

Specific
"Stared at screens for 8 hours, head is buzzing, can't slow my thinking down."

Claude maps "head buzzing" and "can't slow thinking" to high beta and recommends alpha (8-10Hz) with coherent breathing. Protocol is calibrated to the exact arousal state described.

Generic
"I need to focus for a meeting."

Could be 18Hz beta or 40Hz gamma. Without knowing your current arousal state, the choice is arbitrary.

Specific
"About to present to 30 people, feeling scattered, need to be sharp fast."

Claude hears "scattered" plus urgency and recommends 40Hz gamma with box breathing at a short onset duration (8 min), not a 20-minute alpha session that would drop arousal too far before a presentation.

What the AI actually does

Claude reads four signals from your voice.

When you speak into the intake, Claude's analysis reads beyond what you say. The transcript is secondary. The inference is primary.

01
Pace and cadence

Fast speech with truncated phrases signals beta-state overdrive, pointing toward alpha or theta for deceleration. Slow, hesitant speech can indicate delta proximity or low arousal, suggesting a gamma wake-up protocol. The speed and rhythm of what you say carries more information than the words.

02
Stated context

Claude cross-references your words against a band-to-use-case map: sleep issues to delta, focus deficits to gamma, anxiety to alpha or 7.83Hz Schumann, grief or somatic heaviness to theta. If you name your goal clearly, the mapping is direct. If you describe a state without naming it, the inference happens from semantic pattern, not keyword matching.

03
Emotional register

Frustration, flat affect, hyperactivity, and sadness each map to different autonomic states. Claude is not diagnosing. It is reading the nervous system condition implied by the emotional tone and selecting the frequency most likely to shift it. The model was trained on the Mortis band-to-condition matrix, not on generic emotional classification.

04
Time of day

The intake timestamp is factored into the recommendation. A 7am intake suggesting stress will likely receive a different protocol than the same words spoken at 11pm. Morning cortisol biology favors beta and gamma. Evening parasympathetic wind-down favors alpha, theta, and delta. The time context prevents protocols that would physiologically conflict with your circadian position.

No audio is stored. The transcript is processed server-side by Claude and discarded after the session recommendation is returned. No voice biometrics are retained.

After your first session

What to do with what you hear.

The AI will give you a frequency recommendation, a breathwork pairing, and a duration. Here is how to use each one, and what to track.

Frequency
The recommended Hz is your starting point. Trust it for the first session. On session two, you can compare. The personal matrix starts showing your actual response data by session five. Do not change frequency mid-session - the entrainment effect needs 8-12 minutes of consistent carrier to begin taking effect.
Breathwork
The breathwork pairing is for the first 6 minutes of the session. After 6 minutes, let the breath settle naturally. The structured breathing is an onset accelerant, not the session itself. If you find the breathwork distracting, skip it on the first session. Run the frequency alone. Add the breathwork layer once the frequency feels stable.
Duration
The recommended duration is a minimum, not a target. 8 minutes is the floor for measurable entrainment. 20-25 minutes is where most users see their best HRV deltas. If you have 10 minutes, use 10 minutes. A short session with HRV data is more valuable to your matrix than no session.
HRV
If you have an Apple Watch, make sure Mortis has HealthKit access before you begin. The app reads RMSSD before the session and 2 minutes after it ends. That delta is the single most important number the tool produces. No Apple Watch? Log a subjective "helped?" rating after each session instead. It is less precise but still feeds the matrix.
Common first-week patterns

Three patterns that slow your first 14 days. And what to do instead.

These show up in Mortis session data repeatedly. None of them are failures. They are just decisions worth changing.

Pattern 01

Switching frequencies every session.

The matrix cannot calibrate on one session per band. Three consecutive sessions in the same band gives the algorithm enough data to establish a response baseline. After that, variation is fine. Before that, it is just noise.

Do instead: trust the AI recommendation for sessions 1 through 3. Start exploring on session 4.

Pattern 02

Skipping HRV setup because it seems optional.

It is technically optional. But without HRV, you have no feedback signal. The AI can recommend frequencies. It cannot tell you whether they are working for your specific nervous system without the data. Seventy percent of users who skip HRV setup in the first 7 days do not set it up later.

Do instead: open Apple Health, grant HealthKit access now, before your first session. It takes 40 seconds.

Pattern 03

Using earbuds with ANC active during binaural sessions.

Active noise cancellation processes audio in ways that can alter the binaural beat carrier frequency. The two tones need to arrive at your eardrums as-is. AirPods in Transparency mode, or standard over-ear headphones with no processing, are the reliable options.

Do instead: ANC off. Transparency mode on, or passive headphones. Any pair works as long as it is stereo.

If something seems off

The recommendation came back. But it does not feel right. Here is what to do.

The intake is accurate for most people in most contexts. But your context has nuance that Claude may not have captured. These are the four overrides worth knowing.

Override 01

You got a high-frequency recommendation but you are exhausted.

Claude optimizes toward your stated goal. If you said "focus" but your HRV is suppressed, the intake does not always catch that conflict. If you are running on 5 hours of sleep and the intake recommends 40Hz gamma, go to delta (2-4Hz) instead. The body needs recovery before it can benefit from activation.

Signal to check: Apple Health HRV reading from last night. If more than 20% below your 7-day average, run recovery not activation.

Override 02

The frequency and duration feel right but you cannot get still.

If the binaural tone is correct but the ambient noise layer is pulling your attention, switch to None under the ambient layer. The binaural beat works with no additional audio. The ambient layer is purely preference. Some users run sessions in complete silence except for the frequency itself.

Setting path: Session screen > Audio layer > Ambient > None.

Override 03

Your post-session HRV was lower than pre-session.

A single negative delta does not mean the frequency is wrong. HRV varies for many reasons. Run the same band two more times before drawing conclusions. If three consecutive sessions in one band all produce negative deltas, that band is not working for you in your current physiological state. Switch bands and run three more.

Rule of three: no band judgment before three sessions in that band at the same time of day.

Override 04

The intake recommended a band you have already run with no results.

The intake does not know your history until session data is in the matrix. If you ran 10Hz alpha three times in your first week and saw flat HRV, tell Claude that during your next intake. Say "I ran alpha three times, flat delta each time." The intake will recalibrate to a different starting point. The AI uses what you tell it, not what it assumed.

More context in the voice note always produces a better recommendation. Claude can only work with what you say.

The intake's blind spots

Four things the voice intake cannot hear. Tell Claude directly.

The intake reads tone, cadence, and stated state. There are a few categories of context that simply cannot be inferred from a 30-second recording. If any of these apply to your current physiology, speak them explicitly in the intake.

Blind spot 01

Current medications that affect autonomic tone

Beta blockers, SSRIs, stimulants, sleep medications, and some blood pressure drugs directly influence HRV baselines. The intake has no idea you are on any of them. If a medication shifted your resting RMSSD more than 5 ms, the session delta calculations need that context to be interpretable. Say "I am on beta blockers" or "I take sertraline."

Not for dosage adjustment. Only for calibration of expected delta.

Blind spot 02

Recent physical intensity or injury

If you just finished a hard training session, ran a race yesterday, or pulled a muscle two days ago, your HRV is suppressed for reasons unrelated to stress. The intake hears your voice and may assume anxiety or fatigue. Clarify: "I did a hard interval workout this morning" changes the frequency recommendation from activation to recovery.

Training load is the single most common unstated context in the beta cohort.

Blind spot 03

Menstrual cycle phase

HRV varies by roughly 5 to 10 ms across the menstrual cycle for most women, with the lowest values in the luteal phase. The intake has no way to detect this. If you are tracking cycle phase, saying "I am in day 22 of my cycle" or "luteal phase" lets the recommendation account for the expected HRV depression rather than interpret it as acute stress.

Not included in standard HRV baselines. A known unknown that matters a lot.

Blind spot 04

Acute emotional events from the last 24 hours

A fight with a partner, a job loss, a close friend's bad news. You may sound fine in a 30-second recording. Your nervous system is not fine. The intake will often miss this. If something significant happened in the last day, say it: "I just got some hard news yesterday." The recommendation shifts toward recovery and theta, not activation.

The voice analysis picks up some acute stress; it does not pick up dissociation or composed grief.