New Ark Acoustic Detection logoNEW ARKACOUSTIC DETECTION

ACOUSTIC DETECTION · BUYER & TECHNICAL GUIDES / 636

03 / 3THREE DISTINCT QUESTIONS · ONE ACOUSTIC TOPIC

Microphone array supplier for drone detection: Comparing Supplier Test Sheets with Different Measurement Conditions

A microphone array supplier for drone detection may present sensitivity, noise, frequency response, or dynamic-range figures that look directly comparable to another supplier’s figures. They are comparable only after the measurement conditions are aligned. A test sheet can describe a single microphone element, a calibrated element-and-preamplifier channel, or an assembled array with processing. Those are different purchasing objects, even when the row labels look identical.

Discuss PCBA deliveryContact New Ark
WRITTEN FOR636Technical sourcing analyst comparing microphone-array supplier documentation

KEY TAKEAWAY

When a microphone array supplier for drone detection can support a fair comparison

Treat two test sheets as comparable only when each important number has the same definition, reference, unit, frequency basis, weighting, tolerance, and measured configuration. If one of those conditions is missing, the correct procurement result is an unanswered supplier question, not a conclusion that one array is better.

NEW ARKA matching label is not a matching measurement.Keep element, channel, and assembled-array results in separate comparison columns.

01

Microphone array supplier for drone detection: Identical specification names can describe different tests

Technical sourcing analyst comparing microphone-array supplier documentation

The first review task is not to rank the numbers. It is to identify what was actually measured and under what acoustic and electrical conditions.

A supplier’s sensitivity value may be measured with a specified sound pressure, at one reference frequency, in a free field, and at a stated angle. Another sheet may give an open-circuit cartridge value, a system output after a preamplifier, or a nominal production value. All may be called sensitivity. They are not interchangeable until the reference and signal path are known.

The same issue applies to frequency response. An on-axis curve in a free field answers a different question from a diffuse-field curve, an off-axis curve, or a response after array processing. A noise figure may be an A-weighted RMS result, an ITU-weighted quasi-peak result, or an unweighted broadband figure. The label alone does not identify the test.

IEC 60268-4:2018 describes measurement methods and manufacturer characteristics for sound-system microphones, including sensitivity, directional response, dynamic range, and external influences. Its scope is a useful reminder that a specification is tied to a defined measurement method, not merely to a product name.

Start with the measured object

Write the measured object beside every value: capsule, microphone element, element plus preamplifier, complete channel, or assembled array. Also note whether the result is a bench measurement, a factory acceptance result, or a software-derived array metric. This prevents a high element sensitivity from being mistaken for a high-quality assembled-array output.

For a sourcing comparison, a supplier should be able to identify the hardware revision, channel count, signal path, firmware or processing state when relevant, and whether the result came from one sample or a production lot. Without that identity, the number has limited value for an order decision.

02

Align units, frequency references, and weighting conventions

Technical sourcing analyst comparing microphone-array supplier documentation

Small notation differences can change the meaning of a row more than the numerical difference between two suppliers.

Sensitivity may appear as millivolts per pascal or as decibels relative to one volt per pascal. A buyer should preserve the original unit, reference, and whether the value is open-circuit or loaded. Sound pressure values should identify the reference, commonly dB re 20 µPa, and whether the reported quantity is RMS, peak, or peak-to-peak. A number without its reference is not a complete specification.

Frequency references deserve their own column. One sheet may state sensitivity at 250 Hz while another uses 1 kHz. A response limit such as ±2 dB also needs its target curve and frequency interval. A broad nominal range is not equivalent to a guaranteed flat range. Directional microphones may also change response with distance and angle, so an on-axis value cannot stand in for the complete spatial behavior of an array channel.

Weighting conventions are especially important for noise. A-weighted RMS suppresses some low-frequency content and uses an RMS detector; ITU-R BS.468 weighting uses a different filter and quasi-peak method. The figures should not be placed in one ranking column as though they shared a scale.

Make the reference visible

A practical worksheet has separate fields for unit, reference, frequency, sound field, incidence angle, distance, detector, bandwidth, and weighting. If a supplier writes only “noise: 25 dB,” mark the row incomplete rather than converting it or assuming a favorable convention.

The worksheet should also record environmental conditions when the supplier states them. Temperature, humidity, supply voltage, input impedance, and wind or enclosure conditions can affect the result or its interpretation. The goal is not to demand one universal test for every purchase; it is to make differences visible before a comparison is made.

03

Separate typical values from guaranteed limits

Technical sourcing analyst comparing microphone-array supplier documentation

A typical result helps explain what a supplier often observes. A guaranteed limit defines what the buyer may reasonably hold the delivered product to, if the purchasing documents adopt it.

A nominal sensitivity with no tolerance is not the same as a sensitivity range guaranteed for every delivered channel. Likewise, a typical self-noise figure does not establish a maximum noise value, and a representative frequency-response graph does not automatically become an acceptance criterion. The test sheet should say whether a figure is typical, nominal, maximum, minimum, or a production limit.

The distinction matters when an array contains many channels. A single typical element result may be useful for an early technical screen, while channel-to-channel matching, phase consistency, or noise limits may matter for the assembled product. Those requirements need a stated sampling plan and acceptance method rather than a brochure graph.

A supplier can reasonably offer different evidence at different stages. A catalogue value may support a shortlist. A controlled sample report may support evaluation. A purchase specification or certificate may support receiving inspection. These documents should not be treated as equivalent evidence.

Ask what the buyer can enforce

For each preferred number, ask: Is it typical or guaranteed? Does it apply to each channel, each unit, or an average? What sample size and test method support it? Is the stated tolerance measured at final assembly? The answer determines whether the number is a design reference or a contractual acceptance point.

Do not create a performance promise from an attractive typical value. A responsible comparison can say that one sheet reports a lower typical noise figure, while still recording that the sheets do not establish equivalent guaranteed limits.

04

Do not substitute individual microphone data for array data

Technical sourcing analyst comparing microphone-array supplier documentation

An array is more than a group of element datasheets. Its assembled behavior includes mechanics, electronics, timing, channel matching, and any processing that produces the reported output.

Individual microphone data can answer questions about a component: nominal sensitivity, self-noise, frequency response, or overload behavior under the stated test. It cannot by itself establish the assembled array’s direction finding, source separation, beamforming, classification, or localization behavior. Those outputs depend on the complete signal chain and test context.

The physical assembly can also change the acoustic result. A protective cover, support, channel opening, or nearby electronics may affect the sound reaching each element. The processing host may apply gain, filtering, synchronization, calibration, or channel selection. A supplier sheet that reports a finished-array metric should identify whether these factors are included.

Array measurement literature and research infrastructure commonly treat microphone arrays as systems used to locate or analyse sound sources, rather than as a simple sum of independent microphone specifications. For sourcing, that means an array-level claim needs an array-level test description.

Keep three comparison layers

Use three visible layers in the worksheet: element, channel, and assembled array. Put the component datasheet in the first layer, the element-plus-preamplifier or digital-channel result in the second, and system outputs in the third. A blank in the third layer is not a failure of the component; it means the supplier has not provided that system evidence.

For a technical sourcing analyst, this separation also improves quotation scope. The buyer can ask whether the delivered item is an array board, a calibrated module, a finished sensor, or a broader application package without assuming that one supplier’s label matches another’s.

05

Treat unstated conditions as unresolved evidence

Technical sourcing analyst comparing microphone-array supplier documentation

Missing test conditions do not prove poor quality. They prevent a defensible comparison until the supplier clarifies the result.

A buyer should resist filling gaps with industry assumptions. If a sheet does not state whether noise is weighted, whether sensitivity is free-field or pressure response, or whether a frequency curve is measured on-axis, the value can remain in the review as “not comparable yet.” That wording is more accurate than calling the product inferior.

The same restraint applies to graphs. A smooth curve without axis units, tolerance bands, sample identity, or test arrangement may be useful as an illustration, but it is weak evidence for selecting between two assemblies. Ask for the underlying conditions and the intended interpretation before normalizing the graph.

This approach is particularly useful when suppliers use different documentation cultures. One may publish a detailed laboratory report, while another publishes concise nominal specifications. The comparison should reward clarity without pretending that documentation style alone measures acoustic performance.

Use an evidence status, not a forced score

Give each row a status such as comparable, conditionally comparable, supplier clarification required, or not the same measured object. This preserves the reason for uncertainty. A forced score hides whether the problem is a unit conversion, a missing tolerance, or a mismatch between an element and a finished array.

When a supplier answers, append the date, document revision, and exact configuration to the worksheet. That record helps prevent a later brochure revision from silently replacing the evidence used in the sourcing decision.

06

Build a comparison worksheet that produces useful supplier questions

Technical sourcing analyst comparing microphone-array supplier documentation

The best worksheet does more than sort numbers. It shows which missing facts block a fair decision and gives the supplier a precise way to close each gap.

Start with rows for sensitivity, frequency response, self-noise, dynamic range, overload or distortion limit, channel matching, and any array-level output that the purchase actually requires. Beside each row, record the measured object, unit and reference, test frequency or band, sound field, angle and distance, weighting or detector, tolerance, sample basis, and document revision.

Then add a conclusion field that says why the row is comparable or not. For example, two sensitivity values may be conditionally comparable after confirming that both are system outputs at the same reference frequency. Two noise values may remain unresolved because one is A-weighted RMS and the other does not state its weighting. A typical array response graph may remain descriptive rather than an acceptance limit.

For non-operational component sourcing, this discipline is enough to improve an RFQ or technical clarification round. It does not require the buyer to infer a detection range, accuracy, or deployment result from laboratory component data.

Buyer checklist

Before ranking supplier test sheets, ask: What exact object was measured? What are the unit and reference? At what frequency or band was the value obtained? Was the test free-field, diffuse-field, pressure-field, or another stated condition? What angle, distance, bandwidth, detector, and weighting were used? Is the result typical, nominal, or guaranteed? Does it apply to each channel or an assembled array? Which document revision and hardware or software configuration does it cover?

A worksheet that leaves one of these answers blank should carry that blank into the supplier question list. The result is a more honest comparison and a clearer purchase scope, not a forced winner.

QUESTIONS FOR YOUR PROJECT

Buyer checklist for comparing array test sheets

  • Name the measured object: element, channel, module, or assembled array.
  • Record the unit, reference level, RMS or peak convention, and any conversion basis.
  • Confirm the test frequency, frequency band, target curve, and stated tolerance.
  • Identify the sound field, incidence angle, distance, bandwidth, detector, and weighting.
  • Mark every figure as typical, nominal, minimum, maximum, or guaranteed.
  • Check whether the result applies to one sample, every channel, a lot, or an average.
  • Capture the hardware, processing state, document revision, and supplier clarification date.
  • Leave missing conditions as supplier questions instead of assigning an inferred score.

REFERENCES

Sources for this article

Publisher names, dates and relevant contributions are listed below. Confirm offered product details with the supplier.

International Electrotechnical Commission, IEC 60268-4:2018 Sound system equipment — Part 4: Microphones · 2018-09-12
Supports the distinction between defined measurement methods and manufacturer characteristics for sensitivity, directional response, dynamic range, and related microphone specifications.

DPA Microphones, How to read microphone specifications · Accessed 2026-11-20
Supports the explanation that frequency response, sensitivity, sound-field references, distance, nominal values, tolerance, and A-weighted versus ITU-weighted noise figures affect meaningful comparison.

PCB Piezotronics, Microphone Handbook · Accessed 2026-11-20
Supports the distinction between microphone, preamplifier, microphone-system, and array specifications, including the importance of combined system characteristics and stated units or references.

German Aerospace Center (DLR), Aeroacoustic and aerodynamic measurement and analysis methods · Accessed 2026-11-20
Supports the qualitative description of microphone arrays as systems used to locate and analyse sound sources, distinct from individual microphone-element data.

ARTICLE FAQ / 636

Questions on Microphone array supplier for drone detection: Comparing Supplier Test Sheets with Different Measurement Conditions

Practical questions about this article and the proposed product scope.

NEW ARK / ARC-PERIMETERFAQ
RECORD
ACOUSTIC DETECTION LIBRARY

TALK TO NEW ARK

Turn a test-sheet comparison into a clear project scope

New Ark can discuss a project-specific acoustic module or PCBA documentation scope, including which component and assembly evidence should be named in a sourcing inquiry. Final deliverables and acceptance conditions should be agreed for the actual project rather than inferred from a general brochure.

Discuss PCBA deliveryContact New Ark

Hong Kong New Ark Intelligent Technology Limited · Acoustic detection, partner integration and local-assembly discussions.

Back to newsroom