🌿 Effective A-weighted sound pressure level when hearing protectors are worn (selection of hearing protectors) — calculation by the octave-band, HML or SNR method from noise levels and hearing protector attenuation data, PN-EN ISO 4869-2

Environment PN-EN ISO 4869-2

Overview

WHAT THE STANDARD COVERS. Scope from the PKN catalogue card of PN-EN ISO 4869-2:2018-12, in full (our translation of the Polish card text): “This document specifies three methods (the octave-band method, the HML method and the SNR method) of estimating the effective A-weighted sound pressure levels when hearing protectors are worn. The methods are applicable to either the sound pressure level or the equivalent continuous sound pressure level of the noise. Although primarily intended for steady noise exposures, the methods can also be applied to noises containing impulsive components. It is possible that these methods could not be suitable for peak sound pressure level measurements. The octave-band sound attenuation and the H, M, L or SNR values are suitable for establishing sound attenuation criteria for selecting or comparing hearing protectors and/or setting minimum acceptable sound attenuation requirements.” We have the text of the 2018 edition in the preview of ISO 4869-2:2018 (second edition, 2018-10) from the cover to Table 2 in subclause 7.2 (page 5): table of contents, introduction, clauses 1–6 and the beginning of clause 7. The rest — application of the HML method (7.3), the whole SNR method (clause 8) and Annexes A–E — we do not have from the 2018 edition; what we write about it comes from the preview of the first edition, ISO 4869-2:1994 (from the cover to Table B.1 in Annex B), and is marked as such.

ACCORDING TO THE 2018 EDITION (introduction, clauses 1–7.2). The result is an “effective” level, i.e. the A-weighted sound pressure level at the position of the centre of the head with the listener absent, minus the attenuation of the protector — this value is compared with the permissible noise exposure; according to the introduction the effective level is typically 5 dB to 10 dB lower than the level in the ear canal. The introduction calls the octave-band method (which requires the octave-band spectrum of the noise) an “exact” reference method which nevertheless has its own inaccuracies, because it is based on mean values and standard deviations of the attenuation in a group of test subjects, not on the attenuation for the specific wearer; the simpler HML and SNR methods require only the C- and A-weighted sound pressure levels of the noise. The introduction states that, owing to the large spread of attenuation between wearers, all three methods are nearly equivalent in accuracy in most situations, and that in special situations, e.g. with noise dominated by very high or very low frequencies, the HML or the octave-band method has to be used. The result is valid only if the protector is worn correctly and in the same way as by the test subjects according to ISO 4869-1, is properly maintained, and the anatomical characteristics of the test subjects match the wearers; the principal source of inaccuracy is the ISO 4869-1 input data. Note in the introduction: differences of 3 dB or less in the effective sound pressure level for comparable protectors are generally insignificant. Normative references (clause 2): ISO 4869-1 (subjective method for the measurement of sound attenuation), ISO 9612:2009 (determination of occupational noise exposure) and IEC 61672-1 (sound level meters). Clause 4: the one-third-octave-band attenuation of the protector used in the calculations is measured according to ISO 4869-1 — this standard does not measure attenuation. Clause 5: first the protection performance x is selected (the percentage of individuals for whom the achieved protection is equal to or greater than the designated value) together with the corresponding constant α from Table 1: 50 % — 0,00; 75 % — 0,67; 80 % — 0,84; 84 % — 1,00; 90 % — 1,28; 95 % — 1,64; 98 % — 2,00; with α = 1 and x = 84 % the subscripts may be omitted. The assumed protection value in a band is calculated with Formula (1): APVfx = mf − α·sf (mf — mean attenuation, sf — standard deviation, both according to ISO 4869-1). Clause 6 — the octave-band method: it is calculated separately for each noise situation, with Formula (2): L′p,Ax = 10 lg Σ 10^(0,1(Lp,f(k) + Af(k) − APVf(k)x)) dB, summing the bands from 63 Hz to 8000 Hz, where Lp,f(k) is the sound pressure level of the noise in the octave band and Af(k) the A frequency weighting according to IEC 61672-1; if noise or protector data for the 63 Hz band are not available, the summation begins at 125 Hz; the result is rounded to the nearest integer. Subclause 7.2 — H, M and L values: they are calculated from eight reference noise spectra with different (Lp,C − Lp,A) (Table 2, bands 125–8000 Hz, each spectrum normalized to an A-weighted level of 100 dB, differences (Lp,C − Lp,A) from −1,2 dB to 8,4 dB) and from the individual attenuation of each test subject ajf(k): for each subject j and each spectrum i the predicted noise level reduction PNRji = 100 dB − 10 lg Σ 10^(0,1(Lp,Af(k)i − ajf(k))) dB (Formula 15, bands from 125 Hz), from these Hj, Mj, Lj (Formulae 12–14 with the constants di from Table 2), and finally Hx = Hm − α·Hs, Mx = Mm − α·Ms, Lx = Lm − α·Ls — the mean over the test subjects minus α times the standard deviation (Formulae 3–11); H, M and L are rounded to the nearest integer. According to clause 6, in the HML and SNR methods the calculations ALWAYS begin at 125 Hz, regardless of whether data at 63 Hz are available. The table of contents then lists: 7.3 — application of the HML method for estimating the effective level, 8 — the SNR method (8.2 calculation of SNR, 8.3 application), informative Annexes A (example of APV), B (example of the octave-band method), C (example of H, M, L), D (example of SNR) and E (uncertainty of attenuation values and ratings).

APPLICATION OF HML AND SNR — ONLY ACCORDING TO THE 1994 EDITION (7.2, 8.1–8.2). With HML, first PNRx is calculated from H, M, L and the difference (LC − LA) of the noise: for (LC − LA) ≤ 2 dB — PNRx = Mx − (Hx − Mx)/4 · (LC − LA − 2 dB), for (LC − LA) ≥ 2 dB — PNRx = Mx − (Mx − Lx)/8 · (LC − LA − 2 dB); then L′Ax = LA − PNRx, rounded to the nearest integer. With SNR: L′Ax = LC − SNRx; when only the A-weighted level of the noise is available, L′Ax = LA + (LC − LA) − SNRx may be used if the difference (LC − LA) is known (measurement or a table of typical situations). According to the notes of the 1994 edition the unweighted level may be used instead of the C-weighted level — for very low-frequency noise this gives higher L′Ax. Whether the 2018 edition kept these formulae and notes unchanged we have not checked — clauses 7.3 and 8 of the 2018 edition lie outside the text we have read.

WHAT THE 2018 EDITION CHANGED AGAINST 1994 — AND WHAT IT MEANS FOR THE RESULT (comparison of the read parts of both editions). (1) H, M and L: in 1994 they were calculated from the assumed protection values APVfx (mean minus α times the deviation in each band, bands from 63 Hz); in 2018 — from the attenuation of each test subject separately, and only from the resulting H, M, L of the subjects is the mean minus α times the deviation taken, and without the 63 Hz band (Table 2 of the 2018 edition has no 63 Hz column, while the values of the other bands and the constants di are the same in both tables). The same attenuation data of a protector therefore give different H, M and L under the two editions. (2) SNR: in 1994 the sum in Formula (10) covered the bands from 63 Hz (Table 3, pink noise with a C-weighted level of 100 dB); in 2018 the SNR calculations begin at 125 Hz (clause 6) — we do not have the SNR formula of the 2018 edition. (3) Table of α: in 1994 there were values for x = 75, 80, 84, 85 (α = 1,04), 90 and 95 %; in 2018 85 % is gone, and 50 % (0,00) and 98 % (2,00) were added. (4) Impulsive and peak noise: scope of 1994 — the methods are “not suitable” for peak sound pressure level measurements (likewise the PKN card of the 2002 edition: “are not useful”); scope of 2018 — “it is possible that they could not be suitable”. (5) Meters: IEC 61672-1 instead of IEC 651; ISO 9612:2009 added to the references. (6) An informative Annex E on uncertainty was added. (7) According to the 1994 introduction, with noise dominated by high or low frequencies it “may be advantageous” to use the HML or octave-band method, according to the 2018 introduction it is “necessary”. For the octave-band method as such (Formulae 1 and 2) the read texts of both editions show no difference other than the 63 Hz band: in 1994, when protector data for 63 Hz were missing, the 125 Hz values were used; in 2018 the summation then begins at 125 Hz. The H, M, L and SNR values on the packaging or in the instructions of a protector are given by the manufacturer — the read text does not say according to which edition; this has to be checked in the protector data.

STATUS (cards as of 02.10.2026). The card of PN-EN ISO 4869-2:2018-12 (English version) has no “Withdrawn” header — this edition is CURRENT: published 19-12-2018, 32 pages, price group P, Health, Environment and Medicine Sector, KT 21 for Personal Protective Equipment of Workers, ICS 13.340.20, “Introduces: EN ISO 4869-2:2018 [IDT], ISO 4869-2:2018 [IDT]”, “Replaces: PN-EN ISO 4869-2:2002/AC:2007 - Polish version, PN-EN ISO 4869-2:2002 - Polish version”. There is no Polish version of the 2018 edition in the catalogue: the PKN search for “PN-EN ISO 4869-2” (02.10.2026) shows only two items — 2018-12 (English) and 2002 (Polish). Card of PN-EN ISO 4869-2:2002 (Polish version): “Withdrawn and replaced by PN-EN ISO 4869-2:2018-12 - English version”, withdrawal date 19-12-2018; published 17-05-2002, 19 pages; it introduced EN ISO 4869-2:1995 and ISO 4869-2:1994 [IDT] and replaced PN-N-01309-02:1986; additional element — corrigendum PN-EN ISO 4869-2:2002/AC:2007. We have not checked the status of ISO 4869-2:2018 on the ISO website. Legislation: the Regulation of the Minister of Health on tests and measurements of agents harmful to health in the working environment (consolidated text Dz.U. 2025 item 949) names neither PN-EN ISO 4869-2 nor PN-EN 458 — according to its § 16 tests and measurements are carried out with methods from Polish Standards and, where there are none, with methods recommended and validated (whole file of the consolidated text searched: of standards it names only PN-EN 482 and PN-EN 60825-1). The Regulation of the Minister of Economy and Labour of 5 August 2005 on occupational safety and health at work involving exposure to noise or mechanical vibration (Dz.U. 2005 No. 157 item 1318; in the Sejm ELI database status “in force”, no amending acts) does not name either of these standards — it refers to a Polish Standard only for the permissible noise in a rest room; it does, however, define “individual exposure” as the actual exposure level of the worker after taking into account the attenuation obtained by the use of personal protective equipment (§ 2 point 6), the exposure action values — without taking into account the effect of such equipment (§ 2 point 9), and in § 6 para. 2 it requires personal hearing protectors to be selected so as to eliminate the risk of hearing damage or reduce it to the lowest level achievable. The act does not indicate how the attenuation is to be calculated.

HOW MANY LABORATORIES AND IN WHAT FORM (copy of the accreditation scope database, data up to 17.09.2026, read 02.10.2026). The number 4869-2 appears in 20 records of 9 laboratories: AB 327, AB 552, AB 671, AB 752, AB 861, AB 870, AB 940 (4 records), AB 1695 and AB 1749 (2 records each). Current scopes on the PCA website (read 02.10.2026, scope issues from 24.11.2025 to 17.09.2026): each of the 9 laboratories has one item with this number, and in these 9 items the number stands in 11 lines — at AB 861 the item is bilingual (reference in Polish and in English), and AB 870 gives two editions in the same item. Editions in PCA: 2018-12 — AB 327, AB 552, AB 752, AB 861, AB 870, AB 1695, AB 1749; the form “PN-EN ISO 4869-2:2018-2” — AB 940 (so in the PCA document and in the copy); the withdrawn 2002 — AB 671 (alone) and AB 870 (next to 2018-12). Exclusions: AB 552 and AB 1695 — “excluding the HML and SNR methods” (the octave-band method remains), AB 1749 — “excluding the SNR method”; the other six laboratories exclude no method. None of these items is suspended. The “Laboratories” tab searches for entries BEGINNING WITH “PN-EN ISO 4869-2” and shows all 9 laboratories — checked with the tab query on the production server labcoda.pl on 02.10.2026: 9 (under “EN ISO 4869-2” and “ISO 4869-2” — 0).

WHAT IN THESE ITEMS IS A MEASUREMENT AND WHAT A CALCULATION (items in PCA). In all 9 items the object is “Working environment — noise (selection of hearing protectors)” or “Working environment — hearing protectors” (AB 752, AB 870), the measurement — the equivalent sound pressure level in octave bands, in the frequency range (63–8000) Hz, by direct measurement; level ranges — including the equivalent A- and C-weighted levels measured in the same items — from 24 dB (AB 327, AB 861) to 138 dB (AB 752, AB 940), the lower limit from 24 dB to 50 dB (AB 940). Six laboratories (AB 327, AB 671, AB 861, AB 870, AB 940, AB 1749) also measure the equivalent A- and C-weighted sound levels in the same item — the quantities required by the HML and SNR methods; AB 552 and AB 1695, which exclude HML and SNR, give only octave bands; AB 752 gives only octave bands although it excludes no method. The result “from calculation” is named in different ways: “Equivalent A-weighted sound level under hearing protectors” (AB 327, AB 752, AB 861, AB 940, AB 1695, AB 1749), “Effective A-weighted sound level under hearing protectors” (AB 552, AB 671) and “A-weighted sound level under hearing protectors” (AB 870). AB 861 also calculates the “Peak C-weighted sound level under hearing protectors” — with the peak C-weighted sound level measured in the range (69–140) dB.

NEXT TO THIS STANDARD (the same PCA items). PN-EN 458 — all 9 laboratories, in the 2016-06 edition or without the month (“PN-EN 458:2016” — AB 671, AB 752). PN-EN ISO 9612 — 7 laboratories (not AB 752 and AB 870): in the 2011 edition — AB 327, AB 671, AB 861; in the 2025-11 edition — AB 552, AB 940, AB 1695, AB 1749; excluding strategies 2 and 3 (points 10 and 11) — at all except AB 327 and AB 940. At AB 940 the database copy still has PN-EN ISO 9612:2011, while the scope in issue No. 21 of 17.09.2026 already has 2025-11. In-house procedure — AB 327 (ZZH-PB-05, issue 5 of 15.01.2021). No laboratory gives the standard without other documents. Status of PN-EN 458 (cards as of 02.10.2026): card of PN-EN 458:2016-06 (English version) — “Withdrawn and replaced by PN-EN 458:2026-05 - English version”, withdrawal date 28-05-2026; PN-EN 458:2026-05 (English version, published 28-05-2026, 61 pages, introduces EN 458:2025 [IDT]) — no “Withdrawn” header, “Scope”: recommendations for the selection, use, care and maintenance of hearing protectors. All 9 laboratories therefore cite today, next to PN-EN ISO 4869-2, an edition of PN-EN 458 that has already been withdrawn; we have not read the content of any edition of PN-EN 458 and do not state what follows from it for the calculation. Determination of occupational noise exposure according to PN-EN ISO 9612 has a separate entry in the Knowledge Base.

WHERE A SEEMINGLY CORRECT RESULT IS EASY TO GET. In the noise measurement: the octave-band method needs the band levels (from 63 Hz or from 125 Hz), the HML and SNR methods — the C- and A-weighted sound levels; an item with octave bands only provides data for the octave-band method, and whether the C and A levels for the HML and SNR methods may be calculated from the bands instead of measured, the read text of the 2018 edition does not say (the 1994 edition allows the difference (LC − LA) to be taken from measurements or from a table of typical situations). In the protector data: attenuation, APV, H, M, L and SNR come from tests according to ISO 4869-1, not from the laboratory’s measurement; the same name “H, M, L” from data calculated according to the 1994 edition and according to 2018 means different numbers for the same protector (a different way of averaging and the 63 Hz band in 1994) — the read text does not say which edition the manufacturer applies. The choice of protection performance x changes the result: at x = 84 % (α = 1) and at x = 98 % (α = 2) the same data give different APV and different H, M, L — the value of x has to be stated with the result (this concerns subscripts such as H84, SNR84 etc.). In the calculation: L′p,Ax is a calculation whose components are the laboratory’s measurement (Lp,f(k), LA, LC — according to standards on noise measurement, at seven laboratories PN-EN ISO 9612) and the manufacturer’s protector data (mf, sf or H, M, L, SNR), while Af(k) and α are constants from the standards; according to the introduction a difference of up to 3 dB between comparable protectors is generally insignificant. Impulsive and peak noise: according to the 2018 scope the methods may not be suitable for the peak sound pressure level, according to the 1994 edition (and the card of the 2002 edition) they are not suitable for it — the “Peak C-weighted sound level under hearing protectors” at AB 861 stands in an item in which PN-EN 458 and PN-EN ISO 9612 are cited next to PN-EN ISO 4869-2, and the read text does not show which of these documents gives the way of calculating it. Citing the 2002 edition (AB 671, AB 870) means H, M, L and SNR calculated according to the 1994 rules.

WHAT WE DO NOT GIVE. From the 2018 edition we do not have subclause 7.3 (application of HML), clause 8 (calculation and application of SNR, with the reference spectrum) or Annexes A–E (calculation examples, uncertainty) — we give the HML and SNR formulae from the 1994 edition and do not know whether the 2018 edition changed them. We do not give the content of PN-EN 458 (any edition) or of ISO 4869-1, apart from the “Scope” fields of their PKN cards. We do not give the requirements for the test report. We have not checked the regulation on maximum admissible concentrations and intensities of harmful agents (NDN values for noise) with which the exposure is compared.

Method principle

The effective A-weighted sound pressure level under a hearing protector is calculated from the noise measured at the workstation and from the attenuation of the protector determined according to ISO 4869-1. The octave-band method subtracts in each band (63–8000 Hz or 125–8000 Hz) the assumed protection value APV = mean attenuation minus α times the standard deviation, where α corresponds to the selected protection performance x (e.g. α = 1 for x = 84 %), and sums the bands with A weighting. The simplified methods use protector ratings calculated from the attenuation tests: H, M and L — together with the C- and A-weighted sound levels of the noise (HML method), or a single number SNR subtracted from the C-weighted sound level (SNR method); according to the 2018 edition H, M and L are calculated from the attenuation of each test subject separately and without the 63 Hz band.

Applications

• Selection of hearing protectors at the workstation — checking whether the effective A-weighted sound level under the protector is low enough: in the scopes of 9 laboratories, result “from calculation”
• Comparing protectors and setting the minimum required attenuation on the basis of octave-band data or the H, M, L and SNR ratings (scope of the standard)
• Octave-band method only — at AB 552 and AB 1695 (“excluding the HML and SNR methods”); octave bands and HML — at AB 1749 (“excluding the SNR method”)

Key parameters

Standard

Standard number

PN-EN ISO 4869-2:2018-12 (wersja angielska; EN ISO 4869-2:2018, ISO 4869-2:2018 IDT)

Title (PL)

Akustyka — Ochronniki słuchu — Część 2: Szacowanie efektywnych poziomów dźwięku A pod ochronnikami słuchu

Title (EN)

Acoustics — Hearing protectors — Part 2: Estimation of effective A-weighted sound pressure levels when hearing protectors are worn

Step-by-step procedure

1. Noise measurement at the workstation

   Sound pressure levels (or equivalent continuous levels) in octave bands for the octave-band method; C- and A-weighted sound levels for the HML and SNR methods. In the scopes — according to PN-EN ISO 9612 (7 laboratories). PN-EN ISO 4869-2:2018-12 current (card as of 02.10.2026).

2. Protector data and protection performance

   Protector attenuation according to ISO 4869-1 (means mf and deviations sf in the bands) or the H, M, L, SNR ratings; selection of the protection performance x and the constant α from Table 1 (usually 84 %, α = 1).

3. Octave-band method

   APVfx = mf − α·sf in each band; L′p,Ax = 10 lg Σ 10^(0,1(Lp,f(k) + Af(k) − APVf(k)x)) dB from 63 Hz (without 63 Hz data — from 125 Hz) to 8000 Hz; separately for each noise situation; rounding to the nearest integer.

4. HML method

   H, M, L ratings of the protector (according to 2018: from the attenuation of each test subject, bands from 125 Hz); application according to the 1994 edition: PNRx from H, M, L and (LC − LA), L′Ax = LA − PNRx — subclause 7.3 of the 2018 edition is NOT GIVEN.

5. SNR method

   According to the 1994 edition: L′Ax = LC − SNRx (or LA + (LC − LA) − SNRx with a known difference (LC − LA)); in the 2018 edition the SNR calculations begin at 125 Hz, and clause 8 is NOT GIVEN.

6. Result

   Effective A-weighted sound level under the hearing protector (from calculation) stating the method and the protection performance x; comparison with the permissible exposure when selecting the protector — according to documents outside this standard.

Required equipment and apparatus

Health and safety (OHS)

• Noise measurements at workstations with high levels (in the scopes up to 138 dB) — the person measuring wears hearing protectors; the read text of the standard contains no health and safety warnings
• TECHNICAL NOTE: the result under the protector is valid only if the protector is worn correctly and in the same way as in the test according to ISO 4869-1; the methods may not be suitable for the peak sound pressure level (2018 edition), according to the 1994 edition — they are not suitable

Frequently asked questions