# Errata in ISO 13474

> One page of the phonometry documentation, as Markdown. Index: https://docs.jmrp.io/phonometry/llms.txt

Canonical: https://docs.jmrp.io/phonometry/reference/errata/iso-13474/
Language: en
Alternate: https://docs.jmrp.io/phonometry/es/reference/errata/iso-13474/index.md
Version: phonometry 4.0.0
Install: pip install "phonometry[full]"
Author: José Manuel Requena Plens (https://orcid.org/0000-0003-1250-6212)
Published: 2026-10-28
Updated: 2026-11-16T21:36:14Z
License: MIT (https://github.com/jmrplens/phonometry/blob/main/LICENSE)
Summary: Errata in ISO 13474: 3 defects in the printed text, with the evidence and what phonometry does about each.

---

These are the 3 entries of the [errata registry](https://docs.jmrp.io/phonometry/reference/errata/) on ISO 13474:2009, in the order the registry lists them. The registry's introduction says how each claim is checked against the printed page and what the status of an entry means.

## ISO 13474:2009, Annex A.2 (the level of each class credited to Equations (7) and (8))

- **Location:** Annex A (informative), A.2, the paragraph above Table A.3
  that says how its levels were obtained.
- **The print:** "For each octave band, the sound exposure level at location
  A was calculated using Equations (7) and (8). From this, the A-weighted
  sound exposure level for each excess-attenuation class was determined."
- **The problem:** Equations (7) and (8) are the long-term average
  single-event sound exposure level and rating level,
  $10 \lg \left[\sum_{k}\sum_{l} \wp_{\mathrm{atm},k}\,\wp_{\mathrm{exc},l}\,10^{0.1 L_{E,\mathrm{w},k,l}}\right]$
  dB, with $K$ added in the exponent of Equation (8): one number summed over
  every class, from levels already weighted in frequency, with no octave band
  and no class left in it. The paragraph describes a band level for each
  excess-attenuation class and then the A-weighted level of each class drawn
  from it, which are Equation (4), $L_{E,k,l}(j)$, and Equation (5). The
  annex itself uses Equation (7) one step later, for LT1, a single value
  printed on Figure A.3 that it says was "calculated using Equation (7)". The
  cross-reference should read Equations (4) and (5).
- **Evidence:** the paragraph read on PDF page 39 (printed p. 31), Equation
  (4) on PDF page 15 (printed p. 7), Equation (5) on PDF page 16 (printed
  p. 8), Equations (7) and (8) on PDF page 17 (printed p. 9) and the LT1
  sentence on PDF page 42 (printed p. 34), all of BS ISO 13474:2009, the UK
  implementation of ISO 13474:2009 (first edition, 2009-06-15).
- **Library behaviour:** the levels of Table A.3 are taken as printed;
  `frequency_weighted_sel` evaluates Equation (5) and `long_term_sel`
  Equations (7) and (8)
  ([`exposure_distribution.py`](https://github.com/jmrplens/phonometry/blob/main/src/phonometry/environment/assessment/exposure_distribution.py)),
  each under the equation its clause gives it. No change was required.
- **Status:** unreported.

## ISO 13474:2009, Annex A.2 (the shift of Equation (22) printed as 1.04 dB)

- **Location:** Annex A (informative), A.2, the paragraph below Figure A.1 that
  spreads the class density for turbulence.
- **The print:** the paragraph opens with "a normal distribution having a
  standard deviation equal to 5 dB" and continues "In this example, the mean
  value was shifted by an amount, Δμ, equal to 1,04 dB [from Equation (22)]";
  Figure A.3 repeats $\sigma = 5{,}0$ dB.
- **The problem:** Equation (22) is the mean of a lognormal variable and
  evaluates to $\Delta\mu = \sigma^2 \ln 10 / 20$, which is $2.878$ dB at
  $\sigma = 5$ dB. $1.04$ dB is its value at $\sigma = 3$ dB ($1.036$ dB),
  a standard deviation the annex does not use. The rest of the annex was
  computed with $2.878$ dB. The shift keeps the energetic mean of every
  subclass at its centre, which is why the printed LT2 of 37.0 dB agrees with
  the printed LT1 of 37.0 dB; with $\Delta\mu = 1.04$ dB and $\sigma = 5$ dB
  every level of the spread distribution moves up by $1.84$ dB, LT2 reads
  38.8 dB, the peak of Figure A.2 moves from about 30.5 dB to 32.3 dB, and
  $L_{50}$ reads 33.3 dB against the printed 31.5 dB.
- **Evidence:** the paragraph read on PDF page 42 (printed p. 34), Equation
  (22) on PDF page 21 (printed p. 13), Figures A.2 and A.3 on PDF pages 43
  and 44 (printed pp. 35 and 36), all of BS ISO 13474:2009, the UK
  implementation of ISO 13474:2009 (first edition, 2009-06-15). Every value
  was recomputed from Table A.3 on PDF page 40 (printed p. 32) of the same
  document.
- **Library behaviour:** the shift is Equation (22) in closed form,
  [`turbulence_level_shift`](https://github.com/jmrplens/phonometry/blob/main/src/phonometry/environment/assessment/exposure_distribution.py),
  $2.878$ dB at 5 dB, and it is not a parameter a caller can set. The
  conformance check "ISO 13474:2009 Equation (22)" holds it against the
  printed integral evaluated by quadrature, and "ISO 13474:2009 Equation
  (A.4), Figure A.3" holds LT2 at the printed 37.0 dB.
- **Status:** unreported.

## ISO 13474:2009, Annex A.2, Figure A.3 (exceedance levels that are not the roots of Equation (25))

- **Location:** Annex A (informative), Figure A.3, the exceedance levels
  printed beside the cumulative curve.
- **The print:** $L_{95} = 21{,}7$ dB, $L_{50} = 31{,}5$ dB,
  $L_{10} = 40{,}6$ dB, $L_{5} = 43{,}2$ dB and $L_{1} = 48{,}0$ dB, beside
  $\sigma = 5{,}0$ dB.
- **The problem:** Equation (24) defines the probability that the level
  exceeds $x$ as $\int_x^{\infty} \rho^{*}(x')\,\mathrm{d}x'$ and Equation (25)
  the $n$-percent exceedance level as its root. On the distribution of Table
  A.4 spread as A.2 describes, they give 21.6; 31.5; 40.5; 43.0 and 47.5 dB:
  $L_{50}$ agrees and the other four print 0.1 dB to 0.5 dB higher, the more
  so the rarer the level. The annex does not say how it computed them. One
  reading is consistent with the print: a cumulative curve formed from 15 dB,
  where the drawn curves of Figures A.2 and A.3 begin (their axes start at
  10 dB), rather than from minus infinity,
  $1 - \int_{15}^{x} \rho^{*}(x')\,\mathrm{d}x'$; the drawn curve of Figure
  A.3 begins at 15 dB on the line of 1, which is what that reading draws and
  where Equation (24) gives 0.998. Fed the 07:00 to 19:00
  column of Table A.3 as printed, which sums to 1.0004, it gives 21.69;
  31.49; 40.58; 43.16 and 47.97 dB, all five to the printed digit; it then
  adds about 0.17 % to every exceedance, the 0.21 % of the distribution below
  15 dB less the 0.04 % by which the printed column exceeds one, which moves
  the levels at the smallest percentages most. Fed the full-precision
  probabilities that reproduce Table A.4, the same reading gives 21.69;
  31.49; 40.59; 43.17 and 48.05 dB, and $L_{1}$ would print 48.1 dB. The
  reading is therefore a hypothesis and not a reconstruction of the figure;
  what is established is that four of the five printed levels are not the
  roots of Equation (25).
- **Evidence:** Figure A.3 on PDF page 44 (printed p. 36), Figure A.2 on PDF
  page 43 (printed p. 35) and Equations (24) and (25) on PDF page 21 (printed
  p. 13), all of BS ISO 13474:2009, the UK implementation of ISO 13474:2009
  (first edition, 2009-06-15). The levels were recomputed from Table A.3 on
  PDF page 40 (printed p. 32) of the same document.
- **Library behaviour:** `SELDistribution.exceedance` evaluates Equation (24)
  to infinity in closed form and `SELDistribution.exceedance_level` solves
  Equation (25)
  ([`exposure_distribution.py`](https://github.com/jmrplens/phonometry/blob/main/src/phonometry/environment/assessment/exposure_distribution.py)),
  so the example returns 21.6; 31.5; 40.5; 43.0 and 47.5 dB. The conformance
  check "ISO 13474:2009 Equation (25), Figure A.3" holds $L_{50}$ at the
  printed 31.5 dB; the other four printed levels have no conformance check.
- **Status:** unreported.
