¿Qué precisión tiene una app cinta métrica iPhone? Guía profesional
Realistic accuracy numbers for AR measuring apps, what to expect with LiDAR vs without, when to trust the reading, and the five tricks that cut measurement error in half.
Short answer: a modern iPhone with LiDAR will give you measurements within ±5 mm over 3 metres. Without LiDAR, expect ±1-2 cm over the same distance. That's good enough for furniture shopping, real estate listings, and most renovation estimates. It's not good enough for cabinetry or millwork, where you still need a steel rule.
But "accuracy" is a moving target. The same app on the same phone can be off by 5 cm one time and ±2 mm the next, depending on lighting, distance, and how the user holds the phone. Here's what's actually going on inside the AR session and what you can do to consistently get the most precise readings.
What ARKit is actually doing
ARKit (the framework all iPhone measuring apps are built on) does two things at once. It tracks the phone's 6-degree-of-freedom position in space using visual-inertial odometry, the camera reads texture features in your environment 60 times per second, and the IMU (accelerometer + gyroscope) reads motion 1000 times per second. The two signals are fused to estimate where the phone is.
Then, when you tap to place a measurement point, the app casts a ray from the touch location into 3D space and looks for where that ray intersects a detected surface. The intersection point gets a 3D coordinate.
The distance between two points is just the Euclidean distance between their 3D coordinates. Simple in theory. In practice, every step in that chain has error: feature tracking drifts, surface detection is fuzzy, raycast hits can land on the wrong plane.
LiDAR vs no LiDAR: what changes
| iPhone with LiDAR (12 Pro+) | iPhone without LiDAR (Xs, 14) | |
|---|---|---|
| How depth is measured | Direct laser time-of-flight | Inferred from camera parallax + IMU |
| Typical accuracy (≤3 m) | ±5 mm | ±10-20 mm |
| Accuracy at 5 m | ±10 mm | ±30-50 mm |
| Works in pitch darkness | Yes (laser doesn't need ambient light) | No (camera tracking fails) |
| Initialization time | ~1 second | ~3-5 seconds |
Five tricks that cut measurement error in half
- Calibrate by moving the phone first. Before tapping your first point, walk a metre or two around the area you'll be measuring. This gives ARKit enough motion to triangulate depth from parallax, early measurements right after launching the app are the least accurate.
- Use Touch Mode for small dimensions. For anything under 30 cm, AR raycast has a hard time placing the start point exactly where you intend. Touch Mode (in the mode picker) instead places the start point at the phone's physical position. Press the phone against the starting edge, tap once, move to the end, tap again. No raycast = no raycast error.
- Tap a textured surface, not a blank wall. AR plane detection works by finding feature points (corners, edges, dots in texture). A perfectly white plaster wall has zero features. If you need to measure to a blank wall, place a piece of patterned tape or a book where you want to anchor, anything with edges.
- Cross-check long distances by walking. AR drift adds up over distance. For anything over 4 m, measure half-and-half: pick a midpoint, measure from one end to the midpoint, then from the midpoint to the far end. The two halves should add up to within a few millimetres of a single end-to-end measurement. If they don't, the long measurement has drift.
- Don't trust the first reading. Take three. Tap, save, move the phone away, come back, re-tap. Repeat three times. If all three readings agree within ±3 mm, you have a solid number. If they disagree by more than 1 cm, something about the AR session (lighting, tracking quality, surface) isn't favourable, change something and retry.
When NOT to use an AR measure
AR is the wrong tool for:
- Cabinetry, doors, drawer fronts. Anywhere a 1 mm gap matters, a steel rule is faster and more reliable.
- Diagonal across a room over 5 m. Use a laser distance meter, they're cheap, accurate, and don't drift over long distances.
- Outdoors in direct sunlight. LiDAR is washed out by bright IR ambient (sunlight), and camera tracking deals with the same issue from shifting shadows. Better in shade or overcast.
- Reflective or transparent surfaces. Mirrors, polished marble, and glass return bad depth readings. Cover or avoid them.
Apple Measure vs third-party AR apps
Apple's built-in Measure app uses the same ARKit primitives as third-party apps, so raw accuracy is similar. The differences are in features:
- Apple Measure: distance, height (auto-detects a person's height), level. That's it.
- Third-party AR apps (Ruler AR, MagicPlan, RoomScan Pro): the above plus angles, area, polygon measurement, 3D room scans, manual floor plan builders, project folders, PDF export, calibration modes, photo annotations, and so on.
For one-off measurements Apple's app is fine. For anything that needs to be saved, exported, or measured at higher fidelity (Touch Mode, polygon area, accuracy profiles), a dedicated app is more practical.
Bottom line
A modern iPhone AR app is accurate enough to replace a tape measure for most home and pro-light tasks. With LiDAR, ±5 mm is realistic. Without LiDAR, ±1-2 cm is realistic. Use the five tricks above and your readings will improve. For finish work, keep a steel rule in your toolbox.