GPS Drift & Distance Accuracy Analyzer
Upload a single FIT, GPX, or TCX file to inspect raw satellite coordinates vs watch firmware distance, detect pedometer sensor fusion, and analyze sampling rate drops.
Upload Activity Track File
Click or drag your .FIT, .GPX, or .TCX file here for instant accuracy diagnosis
Have you ever finished a run and felt that your smartwatch recorded a distance that was slightly higher or lower than the actual course?
Or perhaps you ran a certified 10K or marathon route, but your sports watch logged 10.25 km or 9.85 km, leaving you wondering whether satellite drift, tall buildings, or watch firmware algorithms were responsible?
The GPS Drift & Distance Accuracy Analyzer is a free, local-first web tool designed to inspect a single activity file (.FIT, .GPX, or .TCX) and uncover exactly how your device computed your distance.
What Causes Smartwatch Distance Discrepancies?#
Smartwatches do not simply measure straight lines between satellite points. Modern running watches (Garmin, Apple Watch, COROS, Suunto) continuously balance multiple sensor inputs:
1. Watch Firmware Sensor Fusion vs. 2D Geometry#
Smartwatches run internal Kalman filtering engines that combine:
- Satellite Lat/Lon Coordinates (2D Haversine position)
- Wrist Accelerometer Motion (Cadence and stride length dead-reckoning)
- Barometric Altimeter Pressure (3D incline hypotenuse vs flat 2D distance)
When satellite reception weakens under thick tree canopies, urban canyons, or overpasses, the watch hardware relies heavily on internal pedometer stride estimation. This often creates a 50m to 300m discrepancy between the watch summary distance and raw satellite coordinates.
2. Smart Recording vs. 1-Hz Sampling#
If your device is set to Smart Recording (saving data points every 3 to 7 seconds to conserve battery) rather than 1-Hz Recording (saving data every 1 second), straight lines drawn between distant points will cut sharp corners on turns, reducing your total recorded distance.
3. Stationary GPS Micro-Drift & Auto-Pause Delays#
When waiting at traffic intersections, GPS receivers experience subtle signal oscillation (1–3 meters per second). If auto-pause is disabled or takes a few seconds to trigger, this stationary jitter accumulates phantom distance over the course of a long run.
100% Private & Browser-Based#
Your GPS location data is completely private. All file parsing, Haversine calculations, and diagnostic analysis take place entirely inside your local web browser. Zero track data is ever uploaded to external cloud servers.