Why can the displayed pattern be misleading?
NI's sampling and aliasing reference shows how an inadequately sampled signal can appear at a false lower frequency. It also describes alias frequency in relation to nearby multiples of the sampling frequency. The displayed trend alone cannot identify the original waveform uniquely.
Commissioning Docs is independent and is not affiliated with NI. This is an evidence interpretation aid, not an instruction to connect an instrument, choose a test setting or bypass an approved procedure.
Sources: NI: NI: Acquiring an Analog Signal, bandwidth, sampling and aliasing, checked 22 October 2026
What does a simple alias example establish?
Illustrative example. Assume a fictional pure 12 Hz sine wave is sampled uniformly at 10 samples per second, with no filtering that removes that component. The difference from the nearest 10 Hz sampling multiple is 2 Hz, so the samples can appear consistent with a 2 Hz sinusoid.
The example explains ambiguity. It does not diagnose a real plant oscillation or show that changing the graph's time axis repairs the data. A saved trace without the necessary original acquisition evidence cannot establish which input generated it.
Which acquisition evidence should stay with the result?
Separate the device's actual acquisition settings from later logging, averaging, export or display settings.
| Evidence | Record | Review question |
|---|---|---|
| Input | Approved signal and operating case | What could be present? |
| Acquisition | Actual rate and instrument mode | What was sampled? |
| Filtering | Applicable bandwidth and filter evidence | What could reach conversion? |
| Processing | Averaging or decimation record | Was data changed later? |
| Trace | Original file and settings revision | Can the result be reconstructed? |
| Decision | Qualified technical review | What remains unresolved? |
What should a commissioning reviewer conclude?
A pattern is an observation until the measurement-chain review supports its interpretation. Preserve the original data and ask the responsible technical reviewer whether the approved method can establish the claimed behavior or whether another approved observation is required.
This question differs from how often a logger saves a row: even a correctly timestamped row can contain an aliased observation. Engineering sign-off is performed only by the registered engineer on the job.
Common questions
What else should the bid team know?
Does the slow trend prove the source was slow?
No. Check the actual acquisition and filtering evidence.
Can display interpolation remove aliasing from already acquired samples?
No. A smoother display does not recover a uniquely known original signal.
Is this an approved test setup?
No. Use the responsible technical team and approved procedure.