Detect control points
The Control Points section opens with the split importer at its top: paste or photograph a split-times list and Dibber places the controls along your timestamped track — see splits, results & comparison. Below it, Detect Controls finds the control circles on your map image automatically. Requirements and good-to-knows:
- Needs an imported and aligned map image — complete map alignment first.
- Detection runs in your browser using a small ML model (~11 MB, downloaded once and cached — later runs work offline). Nothing is uploaded.
- Two modes: fast for typical maps and tiled for large, high-resolution images.
- Control numbers may need fixing — edit any detection by tapping the map to add points manually, adjust, or use “Clear all control points” to start over.
Route classification
Classify matches your GPS track against OpenStreetMap data and sorts every segment by surface and way type (the on-path vs off-grid bar). Controls:
- Snap threshold (default 5 m): how far a track point may be from an OSM way and still match it.
- Short-segment threshold (default 20 m): segments shorter than this get smoothed away.
- Hover a segment to highlight it on the map; edit or delete segments that OSM got wrong.
- Accuracy depends on OpenStreetMap coverage — remote areas classify less completely.
Pace settings
The Pace panel drives every time estimate in Improve and Analyze:
- A base pace plus factors per surface and way type.
- Incline penalty and decline bonus.
- Each value shows its source: default, calculated (derived from this route's data) or user (you set it).
- A reset restores the defaults.