# Sound Insulation Survey Method (ISO 10052)

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

Canonical: https://docs.jmrp.io/phonometry/buildings/insulation/insulation-survey/
Language: en
Alternate: https://docs.jmrp.io/phonometry/es/buildings/insulation/insulation-survey/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: 2026-07-28
Updated: 2026-11-16T21:36:10Z
Standards: ISO 10052:2021
Verified: 3 conformance checks (ISO 10052:2021) https://docs.jmrp.io/phonometry/reference/conformance/#room-building-acoustics
License: MIT (https://github.com/jmrplens/phonometry/blob/main/LICENSE)
Summary: ISO 10052 survey method: octave bands, a hand-held meter and the reverberation index, for airborne, impact, façade and service-equipment quantities.

---

Not every question deserves a full engineering measurement. When a dispute
needs a number this afternoon, ISO 10052 trades accuracy for speed: octave
bands, a hand-held meter and one correction index instead of per-band
reverberation times. This guide covers the survey (control) method: the
reverberation index and its room-class estimate, the airborne, impact,
façade and service-equipment quantities, and the survey fiches. The
engineering-grade methods live in
[Field Insulation Measurement (ISO 16283)](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-field/);
the single-number engines behind the survey ratings in
[Insulation Ratings (ISO 717)](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-ratings/).

## Octave bands and the reverberation index

One number replaces the whole receiving-room correction. Where the
[engineering methods](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-field/)
carry a measured reverberation time per band into every quantity, ISO 10052
collapses that correction into the **reverberation index**
$k = 10\log_{10}(T/T_0)$ ($T_0 = 0.5\ \text{s}$), which may be measured but is
normally read off a table from the room's construction and volume. Every
survey quantity is then an addition or subtraction of $k$:

$$
D_{\mathrm{n}T} = D + k, \qquad
D_\mathrm{n} = D + k + 10\log_{10}\frac{A_0 T_0}{0.16\,V}, \qquad
R' = D + k + 10\log_{10}\frac{S T_0}{0.16\,V},
$$

$$
L'_{\mathrm{n}T} = L_\mathrm{i} - k, \qquad
L'_\mathrm{n} = L_\mathrm{i} - k - 10\log_{10}\frac{A_0 T_0}{0.16\,V}, \qquad
D_{2\mathrm{m,n}T} = D_{2\mathrm{m}} + k,
$$

with $A_0 = 10\ \text{m}^2$ and $V$ the receiving-room volume. The two
normalized quantities carry the same $10\log_{10}(A_0 T_0/(0.16\,V))$ term
with opposite signs, because a level difference improves when the receiving
room is more absorbent and an impact level falls. Where the common partition
area has not been measured, the survey method substitutes $V/7.5$ for $S$ when
that is larger, $V$ being the volume of the receiving room, which should be
the smaller of the pair (Clause 3.6). It is a default partition area inferred
from typical dwelling proportions, and because it can only raise $S$ it can
only raise $R'$ relative to a measured area; state in the report when it was
used, together with the case Clause 3.6 also flags, a common area below
10 m². The clause references follow ISO 10052:2021; the formulas and the
reverberation-index table are identical in the harmonized
EN ISO 10052:2004+A1:2010.

The reverberation index is either **measured** (feed the reverberation time to
`reverberation_index(T)`) or, in a control survey, **estimated** from the room
type and volume with `estimate_reverberation_index(V, room)`: furnished
`"kitchen"` / `"bathroom"` / `"furnished"`, or the unfurnished construction
classes `"a"`–`"h"` and the mixed `"a+e"`…`"d+h"`. Clause 6.5 splits that in
two: **Table 3** classifies the room and gives it its letter, **Table 4**
turns the letter and the volume into the index. A fourth quantity unique to
this method is **service-equipment noise** $L_{XY}$, the energy average of
three A- or C-weighted measurements taken to a fixed positional recipe set out
under [Running the survey](#running-the-survey-clauses-5-and-6).

The letters are not arbitrary: Table 3 is a 2 × 2 × 2 grid over the weight of
the walls and ceiling, the hardness of the floor covering and the weight of
the floor construction.

| Walls and ceiling | Soft covering, light floor | Soft covering, heavy floor | Hard covering, light floor | Hard covering, heavy floor |
| :--- | :---: | :---: | :---: | :---: |
| Light | `a` | `b` | `c` | `d` |
| Heavy | `e` | `f` | `g` | `h` |

A **light wall** is plasterboard or timber studwork, and a heavy wall carrying
a plasterboard lining counts as light; a **heavy wall** is unlined masonry or
block. A **light floor** is planks or boards on timber beams, a **heavy
floor** a concrete slab. The **covering** is carpet (soft) against tiles or
timber flooring (hard). Where the room mixes two constructions over roughly
equal areas, average the two letters: that is what the mixed classes
`"a+e"`…`"d+h"` are; where one construction clearly dominates by area, use its
letter alone. Note that the covering and the floor construction are two
separate axes, so "a hard floor" on its own does not pick a class.

Table 4 comes from a statistical survey of European dwellings built between
1960 and 1980 (NOTE 1): the standard deviation of the tabulated indices is
about 1 dB, and the table is valid for $T_0 = 0.5\ \text{s}$ and rooms up to
150 m³. Changed construction methods or habitation habits can shift it
systematically, so measure $T$ when the result has to hold up.

```python
import numpy as np
from phonometry import building

# Octave-band levels (125-2000 Hz) and the measured receiving-room T.
l1 = np.array([88.0, 90.0, 92.0, 92.0, 90.0])
l2 = np.array([55.0, 51.0, 47.0, 41.0, 35.0])
k = building.reverberation_index([0.70, 0.60, 0.50, 0.45, 0.40])   # k = 10 lg(T/0.5)
res = building.survey_airborne_insulation(l1, l2, k, volume=50.0, area=12.0)
print(np.round(res.d_nt, 1))          # [34.5 39.8 45.  50.5 54. ]  DnT = D + k
print(res.rating.rating, res.rating.c)        # 49 -1  ->  DnT,w (C)
print(res.r_prime_rating.rating)               # 48  ->  R'w

# No reverberation time measured? Classify with Table 3, then read Table 4:
# heavy masonry walls, concrete slab, tiled floor, 35-60 m3 -> class "h".
k_est = building.estimate_reverberation_index(50.0, "h")
print(k_est)                                   # [5.  5.5 6.  5.  5.5]

# Service-equipment noise: one corner measurement then two central ones
# (Clause 6.3.4), each over a separate full operating cycle.
se = building.survey_service_equipment_level(
    [35.0, 30.0, 32.0],          # position 1 (corner), then two at position 2
    reverberation_index=3.0,     # scalar k = 3 dB, i.e. T ~ 1 s
    volume=50.0,
)
print(round(float(se.l_xy), 1), round(float(se.l_xy_nt), 1))   # 32.8 29.8

res.plot()   # DnT vs shifted ISO 717-1 reference (needs matplotlib)
```

![Survey-method airborne insulation: the raw level difference D and the standardized DnT across the five octave bands, with the reverberation-index correction k shaded between them](https://docs.jmrp.io/phonometry/media/survey_insulation.svg)

*The reverberation index $k = 10\log_{10}(T/T_0)$ shifts the raw level difference
$D$ into the standardized $D_{\mathrm{n}T}$: up where the room is live ($T > T_0$),
down where
it is dead. The automatic rating is formed only for exactly 5 octave (or 16
one-third-octave) values.*

**Show the code for this figure**

```python
import matplotlib.pyplot as plt
import numpy as np
from phonometry import building

# Octave-band levels (125-2000 Hz) and the measured receiving-room T.
bands = [125, 250, 500, 1000, 2000]
l1 = np.array([88.0, 90.0, 92.0, 92.0, 90.0])
l2 = np.array([55.0, 51.0, 47.0, 41.0, 35.0])
k = building.reverberation_index([0.70, 0.60, 0.50, 0.45, 0.40])   # k = 10 lg(T/0.5)
res = building.survey_airborne_insulation(l1, l2, k, volume=50.0)

x = np.arange(len(bands))
fig, ax = plt.subplots()
ax.fill_between(x, res.d, res.d_nt, alpha=0.2, label="k = 10 log10(T/T0)")
ax.plot(x, res.d, "--o", label="D (level difference)")
ax.plot(x, res.d_nt, "-s", label="DnT (standardized)")
ax.set_xticks(x, [str(b) for b in bands])
ax.set(xlabel="Frequency [Hz]", ylabel="Level difference [dB]",
       title=f"ISO 10052 survey method: DnT,w = {res.rating.rating} dB")
ax.legend()
plt.show()
```

### `survey_airborne_insulation()` and friends: parameters

| Parameter | Type | Units | Range / default | Notes |
| :--- | :--- | :--- | :--- | :--- |
| `l1` / `l2` | 1D or 2D array | dB | one/band, or `(positions, bands)` | Source / receiving (or outdoor `l1_2m`) levels |
| `li` | 1D or 2D array | dB | one/band, or `(positions, bands)` | Impact levels (energy-averaged over positions) |
| `reverberation_index` | scalar or 1D array | dB | one per band | $k$ from `reverberation_index` or `estimate_reverberation_index`; `survey_service_equipment_level()` also accepts a scalar $k$ |
| `volume` | float | m³ | > 0 | Receiving-room $V$ (for $D_\mathrm{n}$ / $L'_\mathrm{n}$ / $R'$ / normalized) |
| `area` | float | m² | > 0 | Common-partition $S$ (airborne $R'$; $V/7.5$ rule applied) |
| `measurements` | array | dB | exactly 3 | Service-equipment positions (`survey_service_equipment_level`) |
| `room` | str | — | `"kitchen"`/`"bathroom"`/`"furnished"`/`"a"`–`"h"`/`"a+e"`… | `estimate_reverberation_index` room class (Table 3; the index itself is Table 4) |

`survey_airborne_insulation()` returns a `SurveyAirborneResult` (`d`, `d_nt`,
`d_n`, `r_prime`, `rating`, `r_prime_rating`); `survey_impact_insulation()` a
`SurveyImpactResult` (`l_i`, `l_nt`, `l_n`, `rating`);
`survey_facade_insulation()` a `SurveyFacadeResult` (`d_2m`, `d_2m_nt`,
`d_2m_n`, `rating`); `survey_service_equipment_level()` a
`SurveyServiceEquipmentResult` (`l_xy`, `l_xy_nt`, `l_xy_n`).

## Running the survey (Clauses 5 and 6)

Everything above consumes band levels. What separates ISO 10052 from
[ISO 16283](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-field/) is not the
arithmetic, which is simpler, but *how those levels are obtained*: a body
posture, a sweep and a stopwatch instead of five fixed microphone positions
per room. The procedure is short enough to quote in full.

![Plan of a dwelling room with the loudspeaker in the corner opposite the separating element, at least 0.5 m from both walls and facing into the corner, the operator near the centre facing away, and the arm's-length microphone swept through a 180 degree arc four times in about 30 seconds, with an elevation beside it showing the arm rising and falling during each traverse](https://docs.jmrp.io/phonometry/media/diagram_survey_sweep.svg)

**The instrument (Clause 5).** A sound level meter of class 1 or class 2 to
IEC 61672-1, with filters to IEC 61260, adjusted with a calibrator before each
measurement so the readings are absolute levels. The microphone must be a
diffuse-field type; a free-field microphone needs its diffuse-field correction
applied. The tapping machine and the rubber ball must meet ISO 10140-5:2021
Annexes E and F and ISO 16283-2:2020 Annex A. Doors and windows are closed and
shutters normally open throughout (Clause 6.1).

**The source, between rooms (Clause 6.2.2).** Steady noise with a continuous
spectrum; octave-band-filtered noise is allowed, and a broadband spectrum may
be shaped to buy signal-to-noise at high frequencies in the receiving room.
Place the loudspeaker in a **corner of the room opposite the separating
element**, at least 0.5 m from the walls, and if it is a single loudspeaker
system, facing into the corner. Several sources may run at once provided they
are of the same type, driven at the same level by similar but uncorrelated
signals; several loudspeakers inside one enclosure must be driven in phase.
Test a vertical pair from the **lower** room and an unequal horizontal pair
from the **larger** room, unless the other direction was agreed beforehand.

**The sweep (Clause 6.3.1).** Airborne insulation needs the average level in
both rooms, impact insulation only in the receiving room. In each, stand near
the centre of the floor facing *away* from the loudspeaker (source room) or
from the separating element (receiving room), hold the meter out at arm's
length, and move the microphone four times horizontally through 180°, raising
and lowering the arm gently during each traverse. The four rotations take
about **30 s in total**, which is also the integration time. A rotating
microphone on a stand is the accepted alternative: at least 10° to the
horizontal, sweep radius at least 1 m. Without a real-time octave analyser,
repeat the whole sweep once per band and read each 30 s $L_\mathrm{eq}$.

**The tapping machine (Clause 6.2.3).** Near the centre of the floor in the
source room, with the hammer line on the room diagonal. One position is enough
for an isotropic floor and slab. On a ribbed or beamed floor add two more, so
that three positions are randomly distributed over the floor area, with the
hammer line at 45° to the ribs and every position at least 0.5 m from the
edges of the floor.

**The rubber ball (Clause 6.3.2).** The heavy/soft branch measures the
**Fast-weighted maximum** level $L_\mathrm{i,Fmax}$ in octave bands, about 10 s per
measurement, from at least two fixed positions (one near the centre and one
at a different height), separated by more than 0.7 m, at least 0.5 m from any
boundary and at least 1.0 m from the impact position, averaged over positions
band by band. Its single number is the ISO 717-2:2020 Annex D A-weighted sum,
not a shifted reference curve; see
[Heavy and Soft Impact Sources](https://docs.jmrp.io/phonometry/buildings/insulation/heavy-impact-sources/).
Wear hearing protection when measuring in the source room.

**The façade (Clauses 6.2.4 and 6.3.3).** With a loudspeaker, place it outside
at an angle of incidence as close to 45° as possible, preferably on the
ground, with the slant distance $r$ from the source to the centre of the test
specimen at least **7 m** ($d > 5$ m from the façade) and the position chosen
so that the level varies as little as possible over the specimen: under 5 dB
per band across the whole façade (Clause 5). The outdoor microphone sits
$(2.0 \pm 0.2)$ m from the plane of the façade, or further out if that is what
it takes to keep 1 m of clearance from the nearest part of the façade, a
balustrade for instance. Integrate 30 s inside and out. With road traffic
instead, measure the two sides **simultaneously** over 60 s with at least
15 vehicles passing, repeating the indoor sweep through that period; three or
five fixed positions are sometimes necessary.

**Service equipment (Clause 6.3.4).** Two fixed positions, three
measurements. Position 1 is close to the apparently hardest surfaces of the
room, preferably 0.5 m from the walls and from the floor or ceiling, in
practice a corner. Position 2 is in the reverberant field, the central area of
the room. Take **one** measurement at position 1 and **two** at position 2,
each covering one full operating cycle under normal conditions, and each on a
separate cycle. No position may be closer than 1.5 m to a source such as a
ventilation outlet. Dropping the corner position and taking three central ones
is the easiest way to bias this quantity low.

**The band set (Clause 6.4, Table 2).** Airborne insulation and
tapping-machine impact insulation are measured in the five octave bands from
125 Hz to 2000 Hz; the heavy/soft impact source uses 63 Hz to 500 Hz, which is
why 63 Hz appears there and nowhere else. Service-equipment noise is a single
A- or C-weighted level over 63 Hz to 8000 Hz with the stated time weighting.

### Background noise: a floor to meet, not a correction to apply

Unlike the engineering methods, the survey method never corrects for
background noise (Clause 6.2.1). It fixes a signal-to-noise floor instead:
adjust the source so that the receiving-room level exceeds the background by
at least 6 dB in every band, checked by switching the source on and off before
the run. Where the margin still falls below 6 dB, record the uncorrected level
and state in the report that the level difference is underestimated (or, for
service equipment, that the level is overestimated) by an unknown amount.
That is why no correction call appears anywhere on this page:
`building.background_correction` implements the ISO 10140-4 laboratory rule
and must not be applied to survey data. Earlier revisions of this guide said
the opposite, listing the correction as the caller's job; a survey corrected
that way is non-conforming and over-reports the insulation.

The traffic method for façades is the awkward case: the receiving-room
background cannot easily be separated from the traffic signal, so the rule
becomes a duty of care: keep noise from sources inside the building as low as
practicable and say so in the report if you suspect it contaminated the
result, because it can only make the façade look better than it is.

### How much accuracy the survey buys back

The trade this page opens on has a published size. Clause 6.6 NOTE states that
survey results and the corresponding engineering method are estimated to agree
**within ± 2 dB**, and Clause 6.6 itself requires the reproducibility of the
procedure to be checked from time to time in accordance with ISO 12999-1,
particularly after any change of procedure or instrumentation: the same
standard the
[field guide's uncertainty section](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-field/#measurement-uncertainty-iso-12999-1)
runs on. An *estimated* reverberation index adds to that: the tabulated
indices of Table 4 carry a standard deviation of about 1 dB (Table 4 NOTE 1),
and that spread enters every quantity on this page directly, band for band.
Measure $T$ when the result has to survive a dispute; estimate it when the
question is whether a partition is roughly where it should be.

## ISO 10052 survey reports (`.report()`)

The airborne, impact and façade survey results each carry a `.report(path)` that
writes the one-page ISO 10052 survey (control) method field report: the
standard-basis line naming ISO 10052 (octave bands), an optional metadata
header, the octave-band table beside the measured-versus-shifted-reference
curve, the boxed field rating ($D_{\mathrm{n}T,\mathrm{w}}$/$R'_\mathrm{w}$, $L'_{\mathrm{n}T,\mathrm{w}}$ or
$D_{2\mathrm{m,n}T,\mathrm{w}}$), the
survey-method statement, an optional requirement verdict (level differences pass
at or above it, the impact level at or below it) and a footer. The airborne
result reports `quantity="dnt"` (default) or `"r_prime"`; `verbose=True`,
`metadata` and `language="es"` behave as in the fiches above.

```python
import numpy as np

from phonometry import building, ReportMetadata

res.report("DnTw_survey.pdf",
           metadata=ReportMetadata(requirement=40.0))   # DnT,w (C; Ctr)

# Impact and façade surveys reuse the same k; li is the tapping-machine level
# in the room below, l1_2m the level 2 m in front of the façade.
li = np.array([66.0, 64.0, 62.0, 60.0, 55.0])
impact = building.survey_impact_insulation(li, k, volume=50.0)
impact.plot()   # L'nT vs the shifted ISO 717-2 reference (needs matplotlib)
impact.report("LnTw_survey.pdf")                         # L'nT,w (CI)

l1_2m = np.array([76.0, 78.0, 79.0, 79.0, 77.0])
facade = building.survey_facade_insulation(l1_2m, l2, k, volume=40.0)
facade.report("D2mnTw_survey.pdf")                       # D2m,nT,w (C; Ctr)
```

![Survey-method impact insulation: the measured octave-band impact level Li and the standardized L'nT with the reverberation-index correction shaded between them, the L'nT,w rating annotated and a note that a live room lowers the standardized impact level](https://docs.jmrp.io/phonometry/media/survey_impact_insulation.svg)

*The impact survey applies the reverberation index with the opposite sign,
$L'_{\mathrm{n}T} = L_\mathrm{i} - k$: a live receiving room ($T > T_0$) **lowers** the standardized
impact level. As in the airborne case, the automatic rating appears for
exactly 5 octave (or 16 one-third-octave) values.*

**Show the code for this figure**

```python
import matplotlib.pyplot as plt
import numpy as np
from phonometry import building

# Octave-band tapping-machine levels below the floor, and the measured T.
bands = [125, 250, 500, 1000, 2000]
li = np.array([66.0, 64.0, 62.0, 60.0, 55.0])
k = building.reverberation_index([0.70, 0.60, 0.50, 0.45, 0.40])
impact = building.survey_impact_insulation(li, k, volume=50.0)

# One line — L'nT vs the shifted ISO 717-2 reference:
impact.plot()
plt.show()

# By hand, showing the sign flip of the correction:
x = np.arange(len(bands))
fig, ax = plt.subplots()
ax.fill_between(x, impact.l_i, impact.l_nt, alpha=0.2, label="-k = -10 log10(T/T0)")
ax.plot(x, impact.l_i, "--o", label="Li (impact level)")
ax.plot(x, impact.l_nt, "-s", label="L'nT (standardized)")
ax.set_xticks(x, [str(b) for b in bands])
ax.set(xlabel="Frequency [Hz]", ylabel="Impact sound pressure level [dB]",
       title=f"ISO 10052 survey method: L'nT,w = {impact.rating.rating} dB")
ax.legend()
plt.show()
```

**Report:** [Survey airborne ISO 10052 example report (PDF)](https://docs.jmrp.io/phonometry/media/iso10052_airborne_example.pdf): One-page survey-method airborne sound insulation field report between dwellings: the metadata header, the octave-band DnT table beside the measured-versus-shifted-reference curve, the boxed DnT,w (C; Ctr) rating, the survey-method statement and a PASS verdict.

Survey airborne fiche (SurveyAirborneResult.report), DnT,w (C; Ctr).

**Report:** [Survey impact ISO 10052 example report (PDF)](https://docs.jmrp.io/phonometry/media/iso10052_impact_example.pdf): One-page survey-method impact sound insulation field report for a floor under tapping-machine excitation: the metadata header, the octave-band L'nT table beside the measured-versus-shifted-reference curve, the boxed L'nT,w (CI) rating, the survey-method statement and a PASS verdict (a lower impact level is better).

Survey impact fiche (SurveyImpactResult.report), L'nT,w (CI).

**Report:** [Survey façade ISO 10052 example report (PDF)](https://docs.jmrp.io/phonometry/media/iso10052_facade_example.pdf): One-page survey-method façade sound insulation field report: the metadata header, the octave-band D2m,nT table beside the measured-versus-shifted-reference curve, the boxed D2m,nT,w (C; Ctr) rating, the survey-method statement and a PASS verdict.

Survey façade fiche (SurveyFacadeResult.report), D2m,nT,w (C; Ctr).

## What this guide covers

- **Covered.** The ISO 10052:2021 survey method: the reverberation index
  $k = 10\log_{10}(T/T_0)$ measured with `building.reverberation_index` or
  estimated from the Table 3 room classes with
  `building.estimate_reverberation_index`; the survey airborne, impact and
  façade quantities ($D_{\mathrm{n}T}$, $D_\mathrm{n}$, $R'$ with the $V/7.5$ rule, $L'_{\mathrm{n}T}$,
  $L'_\mathrm{n}$, $D_{2\mathrm{m,n}T}$) via `building.survey_airborne_insulation`,
  `building.survey_impact_insulation` and `building.survey_facade_insulation`;
  service-equipment noise $L_{XY}$ from exactly three positions via
  `building.survey_service_equipment_level`; the automatic ISO 717 ratings on
  5-octave or 16-third spectra; and the survey fiches through `.report()`.

- **Not covered.** The survey procedure of [Running the
  survey](#running-the-survey-clauses-5-and-6) is documented here but not
  *checked*: the functions consume band levels wherever they came from, so
  nothing verifies the sweep, the source corner, the tapping-machine positions
  or the one-corner-plus-two-central service-equipment recipe. Nothing verifies
  that the 6 dB signal-to-background floor of Clause 6.2.1 was met either, and
  no correction is applied when it was not: that is the method's own rule, not
  an omission. The ± 2 dB agreement with the engineering method is the
  standard's own estimate (Clause 6.6 NOTE), not something this library models
  or propagates: no survey result carries an uncertainty. When the stakes rise,
  step up to [Field Insulation Measurement (ISO
  16283)](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-field/), whose ISO 12999-1
  machinery does produce one. Annex A's report form and Annex B's operating
  cycles for service equipment are also outside the library.

## See also

- [Field Insulation Measurement (ISO 16283)](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-field/):
  the engineering-grade airborne, impact and façade measurements this method
  approximates, and their ISO 12999-1 uncertainty.
- [Service-Equipment and Activity Noise (ISO 16032)](https://docs.jmrp.io/phonometry/buildings/insulation/service-equipment-noise/):
  the engineering method for the service-equipment sound this survey
  measures quickly: band spectra and a measured reverberation time.
- [Heavy and Soft Impact Sources](https://docs.jmrp.io/phonometry/buildings/insulation/heavy-impact-sources/):
  the rubber ball of Clause 6.3.2, its impact force exposure level and the
  ISO 717-2 Annex D single number the survey heavy/soft branch reports.
- [Insulation Ratings (ISO 717)](https://docs.jmrp.io/phonometry/buildings/insulation/insulation-ratings/):
  the reference-curve engines behind the survey ratings.
- [Room Acoustics](https://docs.jmrp.io/phonometry/buildings/rooms/room-acoustics/): the measured
  reverberation times behind the reverberation index.
- [Levels](https://docs.jmrp.io/phonometry/signals/levels/levels/): the energy averaging behind the
  octave-band levels.
- [Theory](https://docs.jmrp.io/phonometry/reference/theory/rooms-buildings/): the
  reference-curve derivation behind the weighted single-number ratings.
- API reference: [`building.measurement.survey_insulation`](https://docs.jmrp.io/phonometry/reference/api/building/survey-insulation/).

## References

- **International Organization for Standardization. (2020).** [*Acoustics — Rating of sound insulation in buildings and of building elements — Part 1: Airborne sound insulation*](https://www.iso.org/standard/77435.html) (ISO 717-1:2020).
  The airborne and façade single numbers of the survey, rated in octave bands; implemented in the insulation ratings guide.
- **International Organization for Standardization. (2020).** [*Acoustics — Rating of sound insulation in buildings and of building elements — Part 2: Impact sound insulation*](https://www.iso.org/standard/69867.html) (ISO 717-2:2020).
  The impact single number of the survey, rated in octave bands; implemented in the insulation ratings guide.
- **International Organization for Standardization. (2021).** [*Acoustics — Field measurements of airborne and impact sound insulation and of service equipment sound — Survey method*](https://www.iso.org/standard/76560.html) (ISO 10052:2021).
  The field survey method: the reverberation-index correction, the standardized/normalized airborne, impact and façade quantities, and service-equipment noise. The formulas and reverberation-index table are identical in the harmonized EN ISO 10052:2004+A1:2010.
- **Vigran, T. E. (2008).** *Building acoustics*. CRC Press. https://doi.org/10.1201/9781482266016
  The building-acoustics context of the field quantities the survey method approximates. ISBN 978-0-415-42853-8.
