Transient Noise Testing: Sound Energy and Sound Power in the Laboratory Brief

When a laboratory needs to test a latch closing, a mechanism releasing or a brief discharge of air, the instruction “measure noise” needs more detail. Comparing the sound released by one action is different from comparing average sound power over a period of operation. Defining the deliverable before discussing the room and acquisition system can reduce rework caused by discovering an unsuitable test process after construction.

Sound energy level describes an event; sound power level describes a radiation rate

Sound power describes the rate of sound-energy radiation and is expressed in watts (W). Sound energy is the integral of sound power over a stated interval and is expressed in joules (J). Their corresponding levels are both expressed in decibels, but sharing the unit “dB” does not make them interchangeable. [1]

Sound exposure level at one microphone position describes accumulated squared sound pressure there, not the sound energy level of the source. A peak measure cannot substitute for event energy either. [1] Reports need the full quantity name rather than just “single-event noise”.

For example, suppose a mechanism completes one release action and then remains on standby. Including that standby period in the averaging interval may change the average result even though the action itself has not changed. Before comparing designs, decide whether the comparison concerns a single action or a complete operating cycle that includes waiting time. Neither is prescribed here as universally preferable; the choice depends on the evaluation objective.

A standard number alone may still point to the wrong method

The official scope of ISO 3744:2025 excludes short-duration impulsive events. [2] An old request template should therefore not route every brief sound into an ordinary sound-power test under that method.

The scope of ISO 3745:2012 addresses sound-power determination in anechoic and hemi-anechoic rooms and uses sound energy level for impulsive or transient noise. [3] ISO lists the 2012 edition as confirmed in 2026, with a 2017 amendment. [4] This helps screen candidate methods; it does not mean that any impact can automatically be reported under that standard. The applicable product noise test code and the full measurement conditions also need checking.

Define “one action” so both parties mean the same thing

Before requesting quotations, it is useful to specify the start and end of the action and which sounds belong to the event. For a releasing mechanism, does subsequent rebound count? Is resetting tested separately? If different designs are analysed using different event boundaries, it becomes difficult to attribute a difference to the product rather than the analysis.

The recording should also allow the state before and after the sound to be checked. Retaining only one displayed result can make later review difficult if the onset was cut off or the next action was included. Specify the retained recording length separately from the final analysis interval. This article prescribes no universal duration in seconds: the actual event and applicable method determine it.

If the action cannot be repeated consistently, procurement discussions also need to address how data from different microphone positions correspond. Ask the supplier to explain whether the positions capture the same event. If separate runs are needed, how will the comparability of those actions be checked? This task should determine whether more simultaneous channels are needed, rather than a general preference for the largest channel count.

Demonstrate the delivery process with a representative action

Consider including a complete test demonstration with a representative specimen in the proposal review. Follow the process from triggering the action and inspecting the recording to selecting the interval and exporting the result, checking that each step is traceable. Agree the repeat count, handling of abnormal events, frequency range and weighting before testing, rather than selecting only the lowest result afterwards.

The deliverables can retain event identifiers, time records, analysis intervals, the names of the reported quantities and method editions. If both sound energy level and average sound power level are needed, list them separately and explain their respective intervals instead of placing both in an ambiguous “noise value” field.

These are recommendations for defining the task and reviewing procurement, not substitutes for standard provisions. Nor does this automatically require a new laboratory. A suitable room proposal must support the agreed test process and reviewable results; a “transient analysis” item on the equipment list is not enough.

References (checked 26 October 2026)

[1] SIST — ISO 3745:2012 public preview, definitions 3.4 and 3.19–3.22
https://preview.sist.si/sist-preview/45362/5080c9a729fb42b08f433df7ed8129f6/ISO-3745-2012.pdf

[2] ISO — ISO 3744:2025, scope and edition status
https://www.iso.org/standard/80866.html

[3] BSI — BS EN ISO 3745:2012+A1:2017, scope
https://knowledge.bsigroup.com/products/acoustics-determination-of-sound-power-levels-and-sound-energy-levels-of-noise-sources-using-sound-pressure-precision-methods-for-anechoic-rooms-and-hemi-anechoic-rooms

[4] ISO — ISO 3745:2012, confirmation and amendment listing
https://www.iso.org/standard/45362.html