BER Confidence-level Calculator
Convert a BER test duration and observed error count into transmitted bits and statistical confidence against a specified BER limit.
Read the BER confidence testing guideEquivalent BER : 1e-12
Poisson approximation to the binomial error count
Confidence versus measurement time
Each curve assumes a fixed number of measured errors. The marker is your current test result.
Current confidence
Zero errors means an observed BER of 0, not a proven BER of zero.
What the calculator is doing
It first computes N = BPS × T. At the specified BER, the expected error count is λ = N × BERS. The confidence is the probability that a link operating exactly at that BER would produce more than E errors. A high value supports the conclusion that the tested BER is better than the specified limit.
Statistical assumptions
The Poisson rare-event model assumes independent errors and a constant BER. Bursts, loss of pattern synchronisation, traffic-dependent faults and changing environmental conditions can invalidate that model.
Required duration if the final measured error count stays at E