By Ryan

GPX file

Drag and drop a .gpx file here, or choose one:

Exports from most GPS watches, phone apps and mapping software work. Analysed locally — never uploaded.

Terrain (for the time estimate) Good path or track Open hillside / grass Rough or boggy ground Boulder fields / scrambling Winter / soft snow

Fitness (for the time estimate) Very fit — regular hill days Fit — frequent walker Average Occasional walker New to hillwalking

Change these at any time — the estimate recalculates without re-reading the file.

How the ascent figure is calculated

GPS elevation is noisy: even standing still, a recorded track wanders up and down by a few metres, and the vertical error is usually worse than the horizontal one. Naively summing every uphill step in the file counts all of that jitter as climbing and can inflate the ascent figure badly on long tracks with many points.

This tool uses a direction-reversal threshold instead. It follows the elevation trend point by point and only banks a climb (or drop) once the trend has reversed by more than 5 m. Oscillations smaller than the threshold are treated as noise and ignored, while every real rise and fall of more than 5 m is counted in full. The result is close to what a barometric altimeter with mild smoothing would report, and it is a more honest planning figure than the raw sum.

Why ascent figures differ between apps

Load the same GPX into three apps and you will often get three different ascent totals. That is normal, and it comes down to method rather than error in your recording:

• Smoothing choices — every app filters elevation noise differently. A small threshold or window keeps more of the jitter and reports higher ascent; aggressive smoothing trims genuine small rises and reports lower. Neither is wrong — they are different definitions.
• DEM lookup versus recorded elevation — some apps discard the elevations in your file entirely and re-derive them from a terrain model (DEM) using your positions. That removes GPS noise but adds the model's own errors, especially on ridges and in steep corries where a small horizontal error moves you a long way vertically.
• Barometric versus GPS altitude — watches with a pressure sensor usually record smoother, more consistent elevation than phones relying on GPS alone, though pressure drift in changing weather introduces its own slow error.
• Recording interval — a track logged every second has far more points, and far more noise, than one logged every 30 seconds, so the same route can produce different totals from the same device on different settings.

The practical upshot: compare ascent figures only when they were produced the same way, and treat any single figure as an estimate. For route planning, the smoothed figure here plus a sensible margin is a reasonable basis.