Mortis is not a relaxation app. It is a precision nervous system training tool. Here's exactly what it does - and how it differs from everything else on the market.
Nine brainwave bands. 132 individual frequencies. Each with its own scientific overview, session duration recommendations, and use-case index. This is not a playlist. It is a reference library for your brain states.
OpenAI Whisper transcribes your voice. Claude analyzes your emotional and cognitive state. ElevenLabs generates a personalized guided session with a voice tuned to your exact words. No tapping through menus. Just speak.
The voice intake analyzes cadence, word choice, emotional register, and time of day simultaneously. Here's what three different user states produce.
These examples use real Claude analysis patterns. The actual output varies based on the full signal: your history, time of day, previous session data, and the specific words you choose. The mapping above is illustrative of the direction, not the exact output.
The voice intake is not keyword matching. Claude is analyzing multiple simultaneous signals in your voice, some linguistic, some paralinguistic. Here is exactly what is read, weighted, and mapped to the frequency library.
How fast or slow you speak, pauses between words, and the rhythm of your sentences. Rapid speech with few pauses indicates elevated sympathetic tone. Slow, measured speech indicates parasympathetic dominance. This is often more accurate than what you say about how you feel.
What you say you're feeling, what you're facing today, your goal for the session. Claude prioritizes contextual language over binary emotional tags. "I'm anxious about a meeting" tells it something different than "I'm anxious." The situation matters.
Voice quality analysis distinguishes between tension (high phonation frequency, constricted range) and flatness (low energy, monotone delivery). Both can indicate stress but of different types: active anxiety versus depletion/exhaustion. Different protocols are recommended for each.
After 5+ sessions, Claude has your baseline voice signature. Deviation from that baseline is often a stronger signal than the absolute reading. A slight flatness that is normal for you registers differently from flatness that is unusual for you.
The analysis normalizes for regional speech patterns. Cadence and prosody are measured relative to your personal baseline, not against a standard accent profile. A fast speaker from New York and a slow speaker from rural Georgia can both receive the same anxiety reading through different absolute speech rates.
You don't need to say the right words. The intake is designed to work with whatever you naturally say about your state. "I'm kind of meh" is as useful as "I'm experiencing moderate fatigue and elevated cortisol." The AI maps informal language to its physiological correlates.
Environmental noise is filtered. You can record the intake in a coffee shop or a car. What matters is your voice signal relative to ambient noise. Sustained low SNR degrades accuracy; find a quiet corner if you want the most reliable recommendation.
Voice data is processed and discarded. The intake does not store a transcript of what you said. The output is a state classification and a protocol recommendation. No recording of the intake content is retained beyond the current session.
Heart Rate Variability is the gold standard biometric for nervous system state. Mortis reads your HRV via Apple Watch HealthKit before every session and after. You see the delta. Every. Single. Time.
Five sessions is the minimum to establish a meaningful personal baseline. Here is what to do in each one to make those sessions count and how to read the number you see at the end.
The first reading is before caffeine, food, or movement. This is your true resting baseline. Use the morning window. The frequency does not matter yet. You are capturing the number, not testing a protocol.
Keep the time consistent (within 30 minutes of session 1). Try a different band. If you ran 40Hz on day one, run 7.83Hz or 10Hz on day two. You are testing band variation at the same circadian point.
Use the voice intake for the first time on session three, not session one. Having two prior HRV readings gives Claude context for interpreting your voice state against an emerging personal baseline. The recommendation will be more accurate.
If you ran sessions 1-3 in the morning, run session 4 in the afternoon or evening. This adds a circadian dimension to your baseline and helps the matrix understand whether your HRV response varies by time of day. Most users see lower deltas in the evening.
After 5 sessions, your personal matrix shows your HRV baseline, your initial band response ranking, and your time-of-day profile. Every session from this point measures a delta from your baseline rather than from a population average.
The Apple Watch must be worn during the session for HRV readings to appear. The app reads from Apple HealthKit before and after each session. HealthKit access must be granted during onboarding. If you are not seeing readings, check the HealthKit permissions in iPhone Settings.
The number itself is less useful than what it does over time. But here are the three scenarios most users encounter when they see their baseline for the first time.
This range is common in active adults with good sleep and moderate stress. The protocol goal here is optimization and maintenance: keeping the baseline stable through high-stress periods and using gamma for performance-specific gains rather than baseline recovery. You are starting from a strong position.
This is the most common Mortis starting range. It reflects chronic sympathetic activation from work stress, inconsistent sleep, or high training load without adequate recovery. The protocol goal here is baseline elevation: alpha and theta sessions, consistent timing, breathwork emphasis. Users in this range see the largest 30-day HRV improvements.
RMSSD below 35ms (at rest, morning measurement) indicates significant autonomic stress. The protocol goal here is stabilization before optimization: 2Hz delta for sleep, coherent breathing daily, and no gamma sessions until the baseline rises above 40ms. If you are in this range, Mortis can help but you should also review sleep, alcohol, and overtraining.
RMSSD ranges vary by age, fitness level, and measurement method. Morning resting measurement via Apple Watch is the most consistent approach. Do not compare your number to population averages until you have 7 days of consistent readings. The trend matters more than the absolute value.
14 scientifically-validated breathing protocols with visual animation synchronized to your breath rate. HRV coherence tracking in real-time. Layer breathwork over any frequency session - or use standalone.
Every breathwork technique in the library produces a different autonomic effect. The pattern is not decorative. It is a direct input to HRV. Here is how to match the right technique to the right situation.
Produces the largest RMSSD increase of any breathwork technique. The 5.5-second cadence synchronizes the baroreflex cycle with the breath cycle, producing maximum vagal activation. Best paired with 10Hz alpha or 7.83Hz Schumann. Primary citation: Lehrer et al., Applied Psychophysiology and Biofeedback 2003.
Balances sympathetic and parasympathetic tone without strongly biasing either. Used by military operators for maintaining performance under stress. Reduces cortisol without inducing drowsiness. Best paired with 40Hz gamma pre-performance. Effect window: 3-5 minutes to achieve autonomic stabilization.
The extended exhale (8s) and breath hold (7s) produce the strongest vagal activation of any short-duration technique. The CO2 buildup during the hold sensitizes chemoreceptors and amplifies the parasympathetic response on exhale. Most effective for sleep onset. Best paired with 2Hz delta. Andrew Weil popularized this; the mechanism is well-documented in the autonomic literature.
The fastest single-breath technique for acute stress response. One physiological sigh reduces cortisol and anxiety within 15-30 seconds. Mechanism: double inhale re-inflates collapsed alveoli, maximizing oxygen-CO2 exchange; extended exhale dramatically increases vagal tone. Balban et al. (Cell Reports Medicine, 2023) showed this outperforms 5-minute meditation for acute anxiety reduction.
The only activating breathwork technique in the library. Hyperventilation cycles increase adrenaline and temporarily reduce CO2, producing a state of heightened alertness. Not a recovery tool. Best paired with 40Hz gamma for high-intensity performance activation. Mortis includes it at Tier 2 evidence and recommends against use in the evening or for users with cardiovascular concerns.
The frequency and breathwork pairing is the core of every session. Posture and instrument are the additional layers. Each one modulates the session differently. Here is how they work and what they change.
Supine (lying flat) produces the highest post-session RMSSD response in the beta cohort. Seated cross-legged produces the most consistent mid-session EEG coherence in the theta/alpha range. Standing produces the weakest HRV response. The mechanism: vagal tone is strongly influenced by postural pressure on the diaphragm and spinal column position.
The binaural carrier runs underneath the instrument track. The instrument layer serves two functions: it gives the conscious mind a focal anchor during the session (reducing the mental noise that interrupts entrainment) and it modifies the emotional quality of the session. Singing bowls, drone instruments, and sustained strings tend to deepen parasympathetic response. Rhythmic percussion and low-distortion electric tends to support beta/gamma activation.
After session 10, the personal matrix begins tracking posture-specific and instrument-specific HRV response data. The voice intake recommendation includes a posture and instrument suggestion once your matrix has enough data to identify which combinations work best for your physiology. Before session 10, the recommendation uses population-level defaults for your frequency band and stated goal.
From the moment you start to the post-session HRV reading, every element is designed to maximize entrainment depth and physiological impact.
Every session generates six data points. Here is where each one goes and what it feeds into over time.
Fed into: personal baseline calculation, training-day recommendation engine, 7-day rolling average
Fed into: session delta calculation, frequency-band response ranking, personal matrix frequency performance table
Fed into: frequency band usage history, protocol stack recommendation, community band breakdown aggregate
Fed into: circadian pattern profile, time-of-day HRV comparison, morning recommendation engine
Fed into: breathwork-frequency pairing analysis, best breathwork recommendation by band, session quality correlation
Fed into: subjective-vs-objective correlation, sessions flagged as "helped", secondary recommendation signal when HRV data is absent
All six data points are stored anonymously in your personal matrix. Apple Watch HRV data is read from HealthKit on-device and not stored on Mortis servers. Subjective data and session metadata are stored in Supabase linked to your anonymous session ID.
A training log, leaderboard, and accountability system built for people who treat their nervous system like athletes treat their body. Not a mindfulness forum. A training community.
Most meditation apps show streaks and session counts. Mortis tracks the metrics that predict whether the practice is actually changing your physiology. Here is the full stack.
Your 7-day rolling HRV average, updated each morning after a session. A rising baseline means your autonomic nervous system is recovering better. A flat baseline means the protocol needs adjustment. This is the primary outcome metric.
Source: Apple HealthKit RMSSD via Apple Watch
For each frequency band, Mortis tracks your average post-session HRV change. This tells you which band produces the strongest response for your nervous system. After 20 sessions, this data personalizes your default protocol stack.
Requires: 3 sessions per band for statistical minimum
Your HRV baseline and session consistency scored against anonymized aggregate data from all Mortis users in your age and activity cohort. Not a competition. A calibration signal. 80th percentile means your baseline is healthier than 80% of comparable users.
Updated weekly. Cohorts: age bracket + reported activity level
Consecutive days with a completed session of 10 minutes or more. Does not include passive listening. Does not reset for one missed day if 7-day consistency stays above 80%.
A composite of HRV trend direction, session consistency, and protocol accuracy (are you using the band your data says is most effective?). Ranges from 0 to 100. Not shown until day 14 when baselines are meaningful.
Subjective mood ratings. Journal word counts. How long you kept your eyes closed. Whether you "felt" it. Mortis tracks physiology. Everything else is anecdote.
No other meditation app can show you this because no other app tracks what you actually need to see. Week one starts with a single baseline HRV reading. Week four ends with a personalized protocol stack that your nervous system has co-authored.
Your first 5 sessions. The voice intake runs before each one and selects your frequency. You're building your HRV baseline. The app needs 3-5 readings to establish your resting RMSSD - the number every subsequent session is measured against.
You have enough data to see which frequencies produce your highest HRV deltas. The matrix begins populating. Sessions 6-10 start to reveal your band preference - most people show a clear bias toward one or two bands by the end of week two.
You start combining frequencies with the breathwork and posture pairings that work for your body. The protocol builder lets you save stacks. You'll run the same stack 3-4 times and see which version produces the biggest delta. This is where personalization actually happens.
The matrix has enough nodes to show you your practice clusters. HRV improvement becomes statistically reliable at this point. Beta-cohort data shows 8-15% above baseline RMSSD at day 30. The matrix is yours. You co-authored it with your nervous system.
Most meditation apps can't describe this range because they have no data from it. Mortis tracks RMSSD, session timing, and frequency response continuously. By day 90, three things converge that don't exist before day 30.
By session 40, your highest-performing two or three frequencies have appeared enough times to be statistically stable. Claude stops recommending the full range and starts optimizing within your confirmed bands. Sessions become shorter because there is less calibration overhead. The AI is no longer guessing.
Early sessions produce noisy RMSSD data: some days +18ms, some days +2ms. Beta cohort data (n=86) shows the 7-day rolling baseline stabilizing between days 45 and 60 in most users. The variance drops. A rising-and-stable baseline is the clearest single signal that the nervous system is adapting rather than just responding to novelty.
At day 90 most users report selecting their own protocol before the voice intake completes, because they have learned to read their own state accurately enough to pre-select. The AI validates the choice more often than it overrides it. This is the goal: a practice that eventually out-grows its training wheels. The matrix stays and grows. The AI remains available. The dependency doesn't.
Beta cohort data shows ~15% of users see flat or declining RMSSD trends despite consistent sessions. In every case where the user reported the data, the cause was one of three external factors: sleep under 6 hours, alcohol 3+ nights per week, or chronic training load without rest. The matrix flags this and the AI will tell you. Mortis cannot fix a recovery deficit. It can show you exactly where the ceiling is.
The typical meditation app retention curve drops steeply at 30 days and again at 90. Mortis is designed for a different arc. Here is what practitioners with 12-plus months of consistent sessions report.
HRV data is noisy. Frequency preferences are emerging. The AI is still exploring. Most users are doing 3 to 5 sessions per week, trying different bands, discovering that the protocol that works for focus does not work for sleep. This is the right phase to make mistakes.
Expected outcome: HRV baseline rising, noisy week-to-week variation.
The matrix has enough data to confirm your top two or three bands. Sessions start to feel more deliberate because you are no longer exploring. The AI's recommendations align with what you already know works. This is when retention spikes in the Mortis cohort: the tool has become specific to you.
Expected outcome: session consistency up, HRV variance down, protocol clarity.
Practitioners at this stage are using the protocol library in specific functional contexts: pre-competition gamma, post-travel delta recovery, creative work alpha, conflict recovery 7.83Hz Schumann. The matrix is a toolkit now. The recommendation engine surfaces second-tier protocols you have data for but have not used consistently.
Expected outcome: wider frequency range, high per-session HRV delta reliability.
Long-term practitioners report that the session is no longer "something they do." It is part of the morning or pre-sleep infrastructure, as automatic as brushing teeth. The HRV data at this stage shows the most interesting patterns: individual band sensitivity increases, and the nervous system responds to shorter sessions more reliably than in months one through three.
Research analogue: TM practitioner cortisol data at 5+ years. The long-term effect exceeds the short-term one.
Long-term outcome data from Mortis users is not yet available (the product launches April 18). The trajectory above draws from TM longitudinal research (Paul-Labrador et al., 2006) and athletic HRV adaptation literature (Plews et al., 2013). We will publish Mortis's own 12-month cohort data when we have it.
The other apps are relaxation tools. This is a precision nervous system training tool. The difference matters.
| Feature | Mortis | Calm | Headspace | Insight Timer |
|---|---|---|---|---|
| Binaural beat frequency library | ✓ 132 freqs | ✗ | ✗ | ~ limited |
| HRV tracking (before + after) | ✓ Apple Watch | ✗ | ✗ | ✗ |
| Voice intake + AI protocol generation | ✓ | ✗ | ✗ | ✗ |
| Isochronic tones | ✓ | ✗ | ✗ | ~ some |
| Interactive breathwork with HRV feedback | ✓ 14 patterns | ~ basic | ~ basic | ~ some |
| Wim Hof protocol | ✓ | ✗ | ✗ | ~ |
| Session training log + streak tracking | ✓ | ~ basic | ~ basic | ✓ |
| Community leaderboards + HRV comparison | ✓ | ✗ | ✗ | ~ |
| Group challenges | ✓ | ✗ | ✗ | ~ |
| Frequency stacking builder | ✓ | ✗ | ✗ | ✗ |
| Schumann Resonance (7.83Hz) protocol | ✓ | ✗ | ✗ | ~ |
| 40Hz gamma (MIT protocol) | ✓ | ✗ | ✗ | ✗ |
| Athlete-specific protocols | ✓ | ✗ | ✗ | ✗ |
| Personalized AI-generated session audio | ✓ | ✗ | ✗ | ✗ |
| Price (annual) | $144-$192/yr Pro rebates to $0 |
$69.99/yr | $69.99/yr | $59.99/yr |
RMSSD (root mean square of successive RR interval differences) is the HRV metric Mortis tracks. Higher is better for recovery. Lower indicates sympathetic dominance. Here is a practical interpretation guide.
High sympathetic load. Cortisol elevation likely. Chronic stress pattern. Protocol: 2Hz delta or 7.83Hz Schumann, 20+ min sessions, passive (no pre-session stimulants).
Below-average autonomic range for most adults. Improvement is common with consistent practice. Protocol: 10Hz alpha or 6Hz theta, 15 min sessions, coherent breathing (5.5 bpm) stacked.
Healthy autonomic range for sedentary to moderately active adults. All band protocols are appropriate. Focus on consistency and protocol stacking to establish upward trend.
Endurance athlete baseline or high parasympathetic tone. 40Hz gamma pre-performance protocols are the primary value add. Use day-relative HRV variation, not absolute, to guide session selection.
Ranges are general population estimates (Shaffer and Ginsberg 2017, Frontiers in Public Health). Individual baselines vary significantly by age, fitness level, and health status. Mortis tracks your personal baseline established over 5 sessions and measures delta from it, not absolute population comparison.
Most people do not know whether their protocol is working because they are not checking for the right signals at the right times. Here are the three checkpoints the app uses and how to read each one.
After 5 sessions you have enough pre/post RMSSD pairs to calculate a median session delta. If the median is positive (post-session HRV higher than pre), the frequency and protocol are moving your nervous system in the right direction. If it is flat or negative, the band or time of day is wrong.
A single-session HRV delta is acute. The more important signal is whether your 7-day rolling RMSSD average is trending upward compared to your initial 5-session baseline. If it is up by any amount at day 30, the protocol is working at the adaptation level, not just the acute response level.
By day 90, most users have developed intuition about which session they need on a given morning before the app recommends it. Checkpoint 3 is accuracy: does the frequency you selected before the voice intake match the recommendation the AI makes? Beta cohort day-90 self-selection accuracy: 71%.
If you are failing Checkpoint 1 after 5 sessions, the most common fix is time of day: run sessions in the morning on an empty stomach before any stimulant. This gives the recommendation engine the clearest HRV signal and the highest probability of a measurable post-session delta. Protocol adjustments should be made one variable at a time: change the frequency, then the time of day, then the breathwork pairing. Not all three simultaneously.
The voice intake will guide your first session. But if you want to understand the system before you open the app, here is the entry point by starting profile.
Most meditation apps expect constant engagement. Mortis is designed around a specific daily window and then stays out of the way. Here is what the product is doing in the hours and days you are not actively running a session, and what you should expect when you reopen it.
The app reads your Apple Watch HealthKit data on a rolling basis. It does not trigger readings; it pulls the ones your watch is already taking. Overnight, post-meal, during workouts, and at rest. All of that data forms the background against which your session deltas are measured. You never see this collection happen, and the app does not ping you about it.
No notifications. No badges. Background HealthKit read only.
Your rolling 7-day RMSSD baseline is recalculated once per day at 3:00 AM local time. This is when the app updates your "today's HRV sits at X percent of your 7-day average" figure on the home screen. No recalculation happens mid-day, so the reference point stays stable through the morning session window regardless of what happens later.
One job per day at 3 AM. Lightweight. Purely local computation.
The app will not ping you about streaks, days since last session, or missed protocol windows. The beta cohort showed that push notifications about missed sessions produced 2.3 times the churn of silence. We removed them in February 2026. If you want a reminder, you set one yourself in iOS Calendar. The app respects your absence.
The only push notification the app will send is an anomalous HRV drop, and only if you opt in.
One email per week (if you opted in during onboarding). One evidence review, one community data note, one change to the matrix if any frequency tier moved. That is it. No promotional content, no cross-sell, no "we miss you" attempts. If you unsubscribe, the app still works the same way.
Opt-in only. Unsubscribe link in every email. Silent default.
The app does not play audio outside a session. It does not use background audio permissions. It does not interrupt whatever else you are listening to. The only time audio is active is when you explicitly press play on a session. This is a surprisingly rare design choice in the wellness app category.
No session audio in the background ever. No ambient "always on" layer.
The product is designed around a single question: what is the minimum amount of user attention required to produce a measurable autonomic shift? Everything else is noise. The app opens, runs a session, closes. It does not ask you to think about it between sessions. The measurement layer runs silently in HealthKit. The recommendation engine waits until you open it to do anything. This is intentional, and it is the opposite of how almost every other wellness product is built.
Most new users try to use every feature in the first three sessions and end up with noise instead of signal. Here is the short list of things the product does well that you should actively not touch until you have a personal baseline. Ignoring these features in week one is the highest-leverage move a beginner can make.
You do not yet have nodes to put in it. The matrix is beautiful in aggregate and meaningless with three sessions of data. Looking at it too early creates false certainty about what is working and what is not.
Come back at day 30. By then you will have 42 nodes on average and the clusters will show real preference patterns.
Comparing your session count to someone on day 200 produces exactly zero useful information in week one. It also primes competition-driven overuse, which corrodes the protocol. The only baseline that matters right now is yours.
Come back at day 21. By then your session count is meaningful relative to your own starting point.
You can pick any frequency from the 132-entry library. Do not. The voice intake recommendation in week one is based on population priors, not personal data. Your personal data does not exist yet. Overriding the recommendation gives you an uncontrolled variable on top of an already-noisy baseline measurement.
Come back at day 14. After 5 sessions you have enough per-band data to override with intent, not curiosity.
Running a 10Hz alpha session followed by a 40Hz gamma session on the same morning is possible and some advanced users do it. In week one, running two bands makes it impossible to attribute any HRV change to a specific band. One session per window. One variable at a time.
Come back at day 45. Once you know which bands produce a positive delta for you, stacking becomes a sharpening tool instead of a noise generator.
The product has roughly forty separable features. The minimum viable week one stack is three of them: voice intake, one session per day from the recommendation, and Apple Watch HRV measurement. That is the entire required surface for the app to start learning your nervous system. Everything else is optimization on top of a baseline you do not yet have.
Try the 40Hz gamma session. Check your HRV before and after. If it moves - you'll understand why this isn't like anything else.