# Accuracy & Testing: Centimetre-level positioning. Sub-millisecond timing > A dual-band RTK GPS device on the car and a base station at the venue time each crossing of a virtual line to within 1 ms. There are no cables to run and no pavement to cut. The same device is also a data logger for the driver. - Canonical URL: https://siderealtiming.com/accuracy - Last updated: 2026-10 --- ## Scorecard - **Verified Repeatability**: ±1 ms — Consistent spread across independent devices on the same vehicle. - **QA Pass Standard**: 100 consecutive hits per unit — Calibrated multi-marker passes against other systems before delivery. - **Auditable Accuracy**: Every hit — Every crossing is stored with telemetry to audit the accuracy of that individual record. - **Optical Benchmark**: 15.6 ms avg — In our testing, discrepancy range (1.3 to 56.3 ms) observed on commercial optical lasers while Sidereal matched within 1 ms. --- ## The physics: Why personal loggers and phone apps can't give millisecond timing ### Personal loggers and phone apps (1–25 Hz): ±1 to 2 metre drift Phone apps use the phone's built-in GPS, which updates once a second and has limited accuracy. Personal data loggers update 10 to 25 times a second, but without RTK corrections their position still drifts by a metre or two. Most GPS loggers are about 100 times less accurate than Sidereal (about 1 m against 1 cm). At 100 km/h a car covers one metre every 36 ms, so a two-metre drift is 72 ms of timing uncertainty, enough to change podium positions in a tight championship. ### Dual-band RTK GPS: ±1 centimetre precision A stationary base station calculates atmospheric and orbital errors in real time and broadcasts corrections across the venue. Each device then works out the moment it crosses the virtual line, with sub-millisecond precision, and records the whole run as data for the driver. --- ## Verification & testing: How we tested our timing accuracy How a crossing is timed, how it compares with optical lasers, what the audit record holds and where to mount the device. ### How virtual line crossings are calculated (Calculations) Sub-millisecond crossing calculations across the virtual line with recorded signal quality. Instead of relying on a physical beam break or raw GPS snapshot, the system calculates the exact moment the device crosses the virtual timing line. Every crossing records signal precision and satellite lock status on the official record. ### On-track laser benchmark (Track testing) How our timing repeatability compares to commercial optical laser beams. We mounted independent Sidereal devices on the same car, crossing the same line at the same moment. The results: - **Multi-device agreement**: All units agree within ±1 ms. Each crossing is taken at the same centimetre-accurate point on the car, regardless of how the car crosses the line. - **Optical beam benchmark**: Commercial laser beams showed 1.3 ms to 56.3 ms of variance (15.6 ms average), while Sidereal matched within 1 ms. - **Production standard**: Every official device must pass 100 consecutive hits within ±1 ms before delivery, measured against other devices and other sensors. ### Audit trail for protests (Audit trail) What data stewards and timing officials can review if a time is questioned. Every crossing stores key verification details alongside the lap time: - **Fix quality**: Whether the device held an RTK fixed solution at the crossing. - **Estimated fix accuracy**: The location accuracy the GPS solution reported for that crossing. - **Correction age**: How old the RTK corrections were at the crossing. - **Position precision**: Estimated horizontal accuracy at the line. - **Crossing speed**: The car's speed as it crossed the virtual line. ### Device placement and antenna position (Mounting) Mount anywhere on the car - **Solo events (autocross, track sprints)**: The antenna is at the same point on the car when it enters and exits the course, so every lap and split is measured from the same position. - **Wheel-to-wheel racing**: For races judged at the line or on track limits, series scrutineering rules should set the mounting location, just like a traditional transponder, so every car is measured from the same position. --- ## Want to compare it against your current system? We welcome running Sidereal side-by-side with your existing photocell beam or transponder setup at an event so you can see the comparison on your own timing sheets. - Email: hello@siderealtiming.com - Contact form: https://siderealtiming.com/contact