RTK has a reputation for being almost magically precise. Many people hear “centimeter accuracy” and assume that once their receiver shows FIX, they’re locked into 1 cm accuracy no matter where they stand or what hardware they use. Anyone who has spent time in the field knows that reality is a bit more nuanced.
RTK can deliver centimeter‑level accuracy or even below that but only when the conditions, hardware, and correction data all line up. Understanding what affects accuracy helps you avoid frustration and get the best possible results from your setup.
Why this belief is so widespread?
A few things feed the misconception:
- Marketing materials often highlight “<1 cm accuracy” without mentioning the fine print.
- Demo videos are usually filmed in perfect open‑sky environments.
- Many apps show a simple FIX/Float indicator, which hides the complexity behind the scenes.
- New users often assume GNSS works like Wi‑Fi: either it works or it doesn’t.
It’s no surprise that expectations get inflated.
What RTK actually does:
RTK works by comparing the carrier‑phase measurements from your rover to a nearby reference station. When the rover resolves the integer ambiguities, it can compute a very precise relative position.
Under ideal conditions, you can expect:
- 1–2 cm horizontal accuracy
- 2–4 cm vertical accuracy
But “ideal conditions” are not the norm for most fieldwork.
If you’re new to RTK, our Beginner’s Guide to NTRIP is a good place to start.
The real‑world factors that limit accuracy
1. Multipath
Signals bouncing off buildings, vehicles, or even the ground can distort measurements. Multipath is one of the biggest accuracy killers, especially in urban areas.
2. Obstructions
Tree canopy, buildings, and terrain block or weaken satellite signals. Even a few branches can cause intermittent FIX drops.
3. Satellite geometry
It’s not just how many satellites you see — it’s where they are in the sky. Poor geometry (high DOP values) reduces accuracy even with a strong FIX.
4. Atmospheric conditions
Ionospheric activity, humidity, and temperature gradients all affect signal travel time. The farther you are from the base station, the more these errors accumulate.
If you want to compare correction networks in your region, check the NTRIP Provider Directory.
Technical factors that matter just as much
Baseline distance
This is one of the most overlooked factors. Shorter baselines mean more correlated errors and better accuracy.
- 0–10 km: excellent
- 10–30 km: good
- 30–50 km: degraded
- >50 km: unreliable for centimeter accuracy
This is why choosing the closest mountpoint is often more important than choosing the most expensive provider. You can explore nearby free and paid networks on our Free NTRIP Services page.
Receiver quality
Low‑cost receivers can achieve FIX, but they may:
- take longer to fix
- lose FIX more easily
- show more noise in the solution
Antenna placement
A €50 antenna placed well will outperform a €500 antenna placed poorly.
Common mistakes include:
- mounting near metal
- placing the antenna too low
- tilting the antenna
- working too close to buildings
Correction data quality
Not all NTRIP networks are equal. Differences include:
- update rate
- network density
- mountpoint type (VRS, MAC, FKP)
- uptime and redundancy
What this looks like in the field
Open field, short baseline
- Fast FIX
- Stable solution
- 1–2 cm accuracy
Urban canyon
- Frequent FIX/Float switching
- Multipath‑induced jumps
- 5–20 cm deviations are common
Under tree canopy
- FIX may drop entirely
- Accuracy degrades to decimeter level
- Multi‑GNSS helps but doesn’t solve everything
Long baseline (40–50 km)
- Slow or unstable FIX
- Vertical accuracy especially poor
How to actually achieve centimeter accuracy
A few practical habits make a big difference:
- Use a multi‑band receiver for faster, more stable fixes.
- Pick the closest mountpoint, not just the most popular one.
- Mount the antenna in a clear, open‑sky location.
- Use a quality antenna with a proper ground plane.
- Keep an eye on DOP values.
- Avoid reflective surfaces and tree canopy when possible.
If you’re unsure which correction service fits your hardware and region, try our NTRIP Provider Finder.
The bottom line
RTK is incredibly powerful, but it’s not magic. The “1 cm everywhere” idea sets unrealistic expectations and leads to frustration when users encounter real‑world conditions. With the right setup, good correction data, and a bit of awareness about your environment, centimeter accuracy is absolutely achievable — just not guaranteed in every situation.