Apexlog User Manual

日本語 · Home · Privacy Policy

Apexlog records GPS + motion telemetry while you ride or drive, and lets you replay and analyze it. This manual covers how to use the app and — importantly — the open NDJSON data format so you can analyze your exported recordings yourself.

• Recording
• Your recordings (library)
• Playback & analysis
• Exporting
• Data format reference
• Analyzing your data
• Permissions
• Free vs Pro

Recording

1. Mount your phone (or pocket it) and open Apexlog.
2. Tap Record. The status panel shows elapsed time, sample count, current speed and current G.
3. Tap Stop to finish. The run is saved automatically as an .ndjson file.

Each recording captures, from the moment you start (rates are the defaults — see Recording modes below):

• GPS at ~1 Hz — latitude, longitude, altitude, speed, heading, accuracy.
• IMU at ~100 Hz — raw accelerometer, gravity-removed (user) acceleration, and gyroscope.

Recording modes: Sport vs Travel

Open Settings (gear icon) → Recording to pick a mode:

• Sport (default) — the mounted, full-rate mode for track days and rides. You can lower the sensor rates to save battery: IMU Off / 25 / 50 / 100 Hz and GPS High / Balanced / Saver. The IMU drives G-force and lean analysis; turn it off and only the GPS track records.
• Travel — a low-power journey logger for trips, with the phone in your pocket or bag. The IMU is off and a GPS point is written every N metres rather than every second, so nothing is recorded while you stand still and a full day stays small (a 1-hour commute ≈ 100 KB):
  • On foot — walking & trains. A point every ~15 m.
  • Vehicle — car & motorbike. A point every ~50 m.

  A travel recording auto-stops after a safety limit (6/12/24 h, configurable) and posts a notification so you notice — reopen the app and start a new recording to continue after a hotel stop or a meal.

Travel mode and Pro: travel logging is free, but keeping it running with the screen locked or the phone pocketed needs background recording (Pro). Without Pro the app warns you when you start: recording will stop as soon as the screen locks.

Foreground vs background

• Free: recording runs while Apexlog is in the foreground (on screen). Leaving the app or locking the screen stops capture.
• Pro — background recording: capture continues with the screen off or while another app (e.g. your navigation) is in the foreground. This is the intended setup for motorcycles, where the nav app owns the screen. While recording in the background a persistent notification (Android) / status indicator (iOS) is shown, as required by the OS.

Your recordings (library)

The home screen lists your recordings, newest first, with file size and start time. Tap one to open playback.

• Free keeps your latest 3 recordings available; older ones are locked (not deleted) until you unlock Pro.
• Pro keeps an unlimited library.

Delete a recording with the trash icon. Deleting is permanent.

Playback & analysis

Opening a recording shows, top to bottom:

• Map — your track on OpenStreetMap. Map tiles are fetched on demand and cached on the device.
• Vehicle gauge — choose from the menu:
  • Top-down — combined G vector seen from above, plus yaw rate.
  • Rear view — lean angle (and lateral G).
  • Side view — pitch angle (and longitudinal G).
• G-G diagram — the friction circle: longitudinal vs lateral G, showing how hard you braked, accelerated and cornered.
• Channel chart — speed / lean / pitch over a scrubbable timeline.

Use the timeline slider to scrub, the play button to animate, and the speed menu (0.5×–8×) to change playback rate. Lean, pitch and heading are computed by fusing the accelerometer, gyroscope and GPS (a complementary filter), after a device→vehicle calibration derived from the ride itself.

Travel recordings (no IMU) replace the vehicle gauge with a journey gauge: an activity icon that follows your speed, plus a trip odometer (distance covered so far). On foot logs show Break ☕ / Walking 🚶 / Road 🚌 / Train 🚆 / Flight ✈️; Vehicle logs show Stop 🚦 (standing ≤1 min — lights, queues) / Break ☕ (parked longer) / Local road 🚗 (<60 km/h) / Highway 🏎 (faster). The map, speed chart and readouts work as usual.

The on-device analysis uses a calibration that aligns the phone’s arbitrary mounting orientation to the vehicle’s axes. The raw log stores device-frame sensor data — see below if you want to reproduce the vehicle-frame math yourself.

Exporting

Export/share is a Pro feature. Each recording row has an export button and a ⋮ menu with three ways out:

• Export GPX / Export KML — converts the GPS track to a standard map format and opens the share sheet. Your original recording is untouched; the export is a separate file. In Settings → Export you can pick which format the one-tap export button uses and whether to embed elevation.
• Share raw (NDJSON) — sends the raw .ndjson file exactly as described in the data-format reference below; nothing is stripped or transformed.

Opening your track in Google Maps / Earth

• Google My Maps (view in Google Maps): on mymaps.google.com create a map → Import → choose the exported KML (or GPX). The route draws on the map and syncs to the Google Maps app under Saved → Maps.
• Google Earth: earth.google.com → Projects → Import KML file — or just open the KML file on a device with Earth installed.
• GPX also imports into Strava, Garmin Connect, komoot, Ride with GPS and most other GPS tools (as an activity or route, depending on the app).

On a Mac/PC you can also copy the files directly off the device over USB (they live in the app’s Documents directory).

Data format reference

Recordings are UTF-8 NDJSON (newline-delimited JSON): one JSON object per line, one sample per line. The format is open and trivial to parse in any language.

File

apexlog-YYYYMMDD-HHMMSS.ndjson

Named from the local wall-clock start time, e.g. apexlog-20260615-181241.ndjson.

• Line 1 is always the meta line.
• Every following line is one gps or imu sample, in roughly time-generated order.

meta line (first line only)

{"type":"meta","app":"apexlog","ver":"0.2.1","startedAt":"2026-06-15T18:12:41.463902+09:00","device":"iPhone16,2","hz":{"imu":100,"gps":1},"mode":"sport"}

Sample lines

Every sample line carries:

gps line:

{"t":1.000,"s":"gps","lat":35.681236,"lon":139.767125,"alt":3.2,"spd":12.5,"hdg":270.1,"acc":4.0}

imu lines. The three IMU sensors arrive as independent stream events at different instants, so each is written on its own line containing only that sensor’s axes (other fields omitted). A line is one of:

{"t":0.010,"s":"imu","ax":0.12,"ay":-0.03,"az":9.79}
{"t":0.011,"s":"imu","lax":0.05,"lay":-0.01,"laz":0.02}
{"t":0.010,"s":"imu","gx":0.001,"gy":0.000,"gz":-0.002}

Notes:

• Axes are in the device frame (the phone’s own X/Y/Z), exactly as the OS sensor APIs report them — not aligned to the vehicle. Mount orientation is arbitrary; align to the vehicle yourself if needed (see below).
• Numeric precision in the file: lat/lon 7 dp; alt/spd/hdg/acc 2 dp; accel/gyro 5 dp; t 3 dp. Non-finite values are written as 0.
• Streams are not time-aligned. To get accel + gyro at the same instant, interpolate each stream independently over t.
• t comes from a monotonic stopwatch (immune to wall-clock changes); map back to absolute time with meta.startedAt + t.

Analyzing your data

A minimal Python example — load a recording, separate the streams, and compute total G and speed in km/h:

import json, math

meta = None
gps, accel, uaccel, gyro = [], [], [], []

with open("apexlog-20260615-181241.ndjson") as f:
    for line in f:
        o = json.loads(line)
        if o.get("type") == "meta":
            meta = o; continue
        t = o["t"]
        if o["s"] == "gps":
            gps.append(o)
        elif "ax" in o:
            accel.append(o)
        elif "lax" in o:
            uaccel.append(o)
        elif "gx" in o:
            gyro.append(o)

# Total acceleration magnitude in G (gravity included)
G0 = 9.80665
for a in accel:
    a["g"] = math.sqrt(a["ax"]**2 + a["ay"]**2 + a["az"]**2) / G0

# GPS speed in km/h
for p in gps:
    p["kmh"] = p["spd"] * 3.6

print(meta["device"], len(gps), "GPS fixes,", len(accel), "accel samples")

Handy conversions:

• Speed: km/h = spd × 3.6, mph = spd × 2.23694.
• Total G (with gravity): sqrt(ax² + ay² + az²) / 9.80665.
• Linear G (motion only): use lax/lay/laz instead of ax/ay/az.
• Gyro to deg/s: deg/s = rad/s × 180 / π.
• Vehicle frame: estimate “up” from the average of the raw accelerometer (specific force points up at rest), find “forward” by correlating longitudinal GPS acceleration with the horizontal IMU, and take “lateral = up × forward”. This is what the in-app calibration does before computing lean/pitch.

Pandas tip: read with pandas.read_json(path, lines=True), drop the first (meta) row, then split by the presence of columns (ax, lax, gx, lat).

Permissions

• Location (while in use; background is Pro): to record your GPS track. Background access only continues a recording you started.
• Motion & fitness / sensors: accelerometer and gyroscope.
• Notifications (Android): the ongoing “recording” notification the OS requires for background capture.
• Internet: to download map tiles during playback and to process purchases.

You can revoke any permission in system settings; features that depend on it stop until you grant it again.

Free vs Pro

Pro is a one-time purchase — ¥1,000 / US$6.99. No subscription. Restore a previous purchase from the paywall (tap Pro, then Restore).