The choice of sensor decides the success of an acoustic inspection – more than any software. Measuring microphone level in a 95 dB hall fails. Mounting an accelerometer on a vibration-decoupled comfort test sees nothing. This article helps make the right choice.

Three sensor families, three physical principles

Microphone (airborne sound)

Measures pressure fluctuations in air. Classical 1/2" free-field microphones cover 20 Hz – 20 kHz at ±1 dB. Pro: contactless, represents customer perception. Con: extremely sensitive to ambient noise – mostly unusable in loud production halls.

Accelerometer (structure-borne sound)

Piezo element converts mechanical acceleration into voltage. Range typically 0.5 Hz – 10 kHz, top units up to 50 kHz. Pro: very robust against hall noise. Con: contact problem – contact pressure and mount position influence the result.

Laser vibrometer

Measures velocity via Doppler effect – contactless, with µm/s resolution. Range up to > 1 MHz. Pro: ideal for light or hot parts, no mass loading. Con: expensive (10–80 k€), alignment-sensitive, requires optical access.

Comparison matrix

CriterionMicrophoneAccelerometerLaser vibrometer
Frequency range20 Hz – 20 kHz0.5 Hz – 50 kHz0.1 Hz – > 1 MHz
Noise immunity★☆☆☆☆ (very low)★★★★★ (very robust)★★★☆☆ (medium)
Mountingcontactlessglued / screwed / magneticcontactless, optical
Part reactionnonemass loading possiblenone
Inline capabilityonly in noise enclosureyes, robustyes, with robot tracking
Cost / sensor200–2,000 €150–1,500 €10,000–80,000 €
Customer perception★★★★★★★☆☆☆★★☆☆☆

Decision guide

Comfort & sound quality (customer ear): microphone is mandatory – only in a quiet environment (hemi-anechoic room).

Inline EOL in noisy halls: accelerometer on the part or test rig. Hall noise is not "heard", inspection robust against neighbouring lines.

Resonance analysis on small or hot parts: laser vibrometer to avoid mass loading.

High-frequency signals (> 20 kHz): acoustic emission or laser. Standard microphones drop out.

Hybrid setups in practice

In many RTE projects we deliberately combine two sensor principles: an accelerometer secures the robust OK/NOK decision, a microphone validates against the originally subjective assessment. This combines inline robustness with customer relevance without requiring an acoustic chamber on the line.

Conclusion

The "best sensor" does not exist – only the right one. For every inspection task we recommend a short feasibility study with two or three sensor types in parallel. If you are unsure what works on your line, request a free initial analysis.