In specifications the terms appear interchangeably: "noise test", "NVH test", "vibration analysis". Technically they denote different strategies with different purposes. Confusing them inspects the wrong phenomenon.
Noise testing
Measures airborne sound radiated by the part. The microphone hears what the customer will hear. Metrics: SPL, A-weighting, psychoacoustic parameters (loudness, sharpness).
Vibration analysis
Measures structure-borne sound – accelerations, velocities, displacements on the surface. Delivers the mechanical excitation pattern: orders, bearing patterns, structural modes.
Crucial difference
Airborne sound is the end of a long transmission chain: mechanical excitation → structure vibration → radiation. Each stage filters. A strong 800 Hz structure mode may be clearly measurable in the part – but if it does not radiate efficiently the customer hears nothing.
Comparison matrix
| Criterion | Noise testing | Vibration analysis |
|---|---|---|
| Customer perception | ★★★★★ | ★★☆☆☆ |
| Root-cause analysis | ★★☆☆☆ | ★★★★★ |
| Hall noise immunity | ★☆☆☆☆ | ★★★★★ |
| Inline-capable w/o enclosure | no | yes |
| Direct standard evaluation (ISO 532) | yes | indirect |
| Investment | high (enclosure) | low |
When does vibration alone suffice?
For inline EOL where OK/NOK matters and correlation to customer perception is validated upfront. No acoustic chamber needed.
When do you need airborne?
Comfort audits, product development, complaint analysis – the subjective impression matters.
Hybrid: model instead of microphone
One-time measurement of the transfer function between accelerometer and microphone in the lab. Inline only structure-borne is captured – airborne is computed virtually. Customer-relevant statement without enclosure on the line.
