- What speed threshold should I use to detect breaks?
- 0.5 m/s (roughly 1.8 km/h) is a sensible default. Anything below that is essentially stationary for cycling or running. Drop to 0.2 m/s for indoor trainer or treadmill workouts where you are truly stationary (or where any speed reading is sensor noise). Raise to 1.0 m/s for trail rides with hike-a-bike sections you want to keep as part of the activity rather than removing as breaks.
- How does this differ from Strava auto-pause?
- Strava auto-pause uses its own threshold logic to compute "moving time" but it does not modify the underlying file - if you export the GPX from Strava, all the break time is still in the trackpoints. This tool actually removes the break trackpoints from the file, so the result is a shorter activity in real terms - total elapsed time drops to match moving time. Use this when you want the file itself to reflect just the moving portions.
- Does break detection work without a speed stream?
- Yes. If the file does not carry a speed field (some older devices and some converted files), the tool falls back to GPS-derived speed - distance between consecutive trackpoints divided by the time delta. It is slightly noisier than a real speed sensor reading but accurate enough for break detection. For indoor trainer rides where there is no GPS, you need a speed stream; otherwise breaks cannot be distinguished from active recording.
- Will my Strava moving time match the cleaned file?
- Close, not identical. Strava's auto-pause uses its own internal threshold (around 0.5 m/s with a few seconds of grace) and applies it on import. After running the break remover with similar thresholds, the cleaned file's total elapsed time will roughly equal Strava's reported moving time of the original. Re-upload the cleaned file and Strava will report a total time close to its previous moving time, with very little difference between elapsed and moving since the breaks are gone.
- What is the minimum-gap duration for?
- It prevents the tool from cutting out every brief slowdown at a turn or intersection. A red light counts as a break only if you sat there for the full minimum duration; a few seconds slow through a corner does not. Default is 30-60 seconds. Lower it to catch shorter pauses (useful for criterium-style rides with many brief stops). Raise it to keep moderate slowdowns intact and only cut genuine multi-minute breaks.
- How is this different from the Section Remover?
- The break remover is automatic - it scans the entire file for stationary periods and removes them all in one pass. The [FIT Section Remover](/tools/fit-section-remover) is manual - you pick exactly one range to cut by dragging handles on the elevation chart. Use the break remover for cleaning up a ride with multiple coffee stops; use the section remover for surgical cuts of specific named segments.
- Will this affect my average speed and pace numbers?
- Yes, in the direction you would expect. After cleaning, average speed equals total distance divided by what was previously moving time only, because the break time is gone from total elapsed time too. Average pace tightens. Average HR and power may also shift because the low-HR low-power break periods are no longer pulling down the averages. The cleaned file generally reads as a slightly harder activity than the original because the rest periods are gone.
- Is the break remover free?
- Yes - free, no signup, no upload. The detection and removal run entirely in your browser. Drop, set thresholds, preview, download.