emission.free_field_qualification

La referencia de la API se publica en inglés en los dos idiomas: se genera a partir de los docstrings del código, que son su texto original.

Qualification of an anechoic or hemi-anechoic room by the inverse square law: ISO 26101:2017, with the criteria of ISO 3745:2012 Annex A as Amendment 1:2017 rewrote it.

A free field is judged by how the level falls away from a small source. Along each straight microphone traverse the measured level $L_{p i}$ at the distance $r_{i}$ from the mathematical origin of the traverse is compared with the inverse square law

L_{p} (r_{i}) = b - 20 l g \frac{r _{i}}{r _{0}} dB, r_{0} = 1 m (ISO 26101 Formula (2))

and the deviation at every point is

Δ L_{p i} = L_{p i} - L_{p} (r_{i}) (ISO 26101 Formula (4))

where $L_{p i}$ has been corrected for the stability of the source with the monitor microphone, $L_{p i} = L_{p i}^{'} - L_{p, ref, i} + L_{p, ref, 0}$ (Formula (1)). The room is qualified out to the largest distance from the origin at which every deviation, on every traverse and at every frequency, stays inside the limits of Table A.1 (the same table in ISO 26101 Annex A and in the amended ISO 3745 Annex A): $\pm 1, 5$ , $\pm 1, 0$ and $\pm 1, 5$ dB in an anechoic room for the one-third octave bands up to 630 Hz, from 800 Hz to 5 000 Hz and from 6 300 Hz up, and $\pm 2, 5$ , $\pm 2, 0$ and $\pm 3, 0$ dB in a hemi-anechoic one.

The fit of b. Formula (2) leaves the source strength $b$ free: it “is adjusted to optimize the fit of the measured sound pressure levels into the tolerance range, to maximize the qualified distance from the test sound source”. Note 1 offers the mean of $L_{p i} + 20 l g (r_{i} / r_{0})$ as a starting value for an iterative search (Formula (3)); no source prints the search. This module does not iterate, because the problem it poses has an exact answer. Write $y_{i} = L_{p i} + 20 l g (r_{i} / r_{0})$ , so that $Δ L_{p i} = y_{i} - b$ . The points out to a distance $R$ fit inside $\pm t$ for some $b$ exactly when the spread of their $y_{i}$ is at most $2 t$ , and the spread over the points nearer than $R$ can only grow with $R$ . So, for one traverse and one frequency, the qualified distance is the distance of the last point of the longest run from the origin whose $y$ spread stays within $2 t$ , and every $b$ in $[max y - t, min y + t]$ over that run qualifies it. Of those the module takes the midpoint, $b = (max y + min y) /2$ , the one that leaves the same margin to both limits and so the smallest largest deviation. A qualified distance cannot be extended by any other $b$ , and no search can find a longer one.

The mathematical origin. ISO 26101:2017 5.1.3.2 and the amended ISO 3745 A.3.3 require every traverse to share one mathematical origin, and that origin to lie within the physical volume the test source occupies; the 2012 annex had instead fitted a collinear acoustic-centre offset per traverse. Given a fixed origin, the distances follow from the measured positions. Given the box the source occupies, the origin is searched inside it for the largest radius of A.2.4, the smallest qualified distance over every traverse and every frequency, since that is the distance Formula (2) asks $b$ to maximize and A.2.4 defines from the origin. A qualified distance ends either where a deviation leaves Table A.1, which the room decides, or at the last point measured, which the measurement decides; moving the origin inside the box moves the distance to that last point by as much as the origin moves, which says nothing about the room. So the search ranks origins by that smallest distance with every run that reaches its last point counted at the distance of that point from the centre of the box, the same for every origin: radii then differ only where the room ends a run, and radii within the 1 mm to which the search resolves the origin count as equal. Ties are broken by how much of the Table A.1 band the curves fill on average (the mean over every traverse and frequency of the half spread of $y_{i}$ as a share of its limit), then by the nearer point to the centre of the box; the mean, rather than the worst band, keeps one band from dragging the origin to the edge of the source at the others’ expense when nothing else separates two origins. In an exact inverse-square field every run reaches its last point, the curves are flattest seen from the point the field diverges from, and the search returns that point. The radius reported is always the distance from the origin returned. The search is a regular lattice of five points per axis followed by a compass search that halves its step down to 1 mm; it is deterministic, stays inside the box by construction and never returns a worse origin than the best lattice point.

What the verdict judges. check_free_field holds the qualification to the amended ISO 3745 Annex A and to the clauses of ISO 26101 it defers to: the deviations of A.2.2 and the radius of A.2.4; the frequencies of A.2.3 (100 Hz to 10 000 Hz, one-third octave bands below 125 Hz and above 4 000 Hz and octave mid-band frequencies between); five to eight traverses (A.3.3), towards the five targets a) to e) of A.3.3 and in the working area, within the angular limits of the source directionality in a hemi-anechoic room; at least 10 points on each traverse and 50 in total within the radius, equally spaced at each frequency, and spaced at most a tenth of a wavelength below 250 Hz and 100 mm above (A.4.3); the spatial resolution of ISO 26101 A.4.3 and the traverse length of ISO 26101 5.1.4.3, which A.2.4 cites; the 6 dB above background of ISO 26101 5.1.2.2 c); the directionality of the test source (ISO 26101 Annex B, verify_source_directionality); and, in a hemi-anechoic room, the reflecting plane of A.2.5. A requirement whose data were not given is listed in FreeFieldCheck.not_judged, and the verdict does not pass until it is judged.

Two readings are this module’s, since no source settles them. The spacing rules of A.4.3 name a frequency below which a tenth of a wavelength applies and above which a fixed length does (250 Hz and 100 mm in the amended ISO 3745, 1 kHz and 25 mm in ISO 26101), and neither says which holds at the edge itself; the one-third octave band that contains the edge frequency is held to the stricter of the two. And “equally spaced” comes with no tolerance: the points within the radius are equally spaced when, walking out from the first, every next point of the grid lies within a tenth of the median gap of where one step would put it. Points between two of them are the additional measurements the last paragraph of A.4.3 recommends near a peak deviation, and do not break the grid.

Auto-generated from the source docstrings by scripts/generate_api_docs.py (make api-docs). Do not edit by hand.

check_free_field

check_free_field(
    results: InverseSquareLawResult | Sequence[InverseSquareLawResult],
    *,
    bandwidth: QualificationBandwidth,
    source_directionality: SourceDirectionalityResult | Sequence[SourceDirectionalityResult | None] | None = None,
    measurement_radius_m: float | None = None,
    reflecting_plane_absorption_coefficient: float | None = None,
    reflecting_plane_margin_m: float | None = None,
    paths_in_working_area: bool | None = None,
    speed_of_sound: float = 343.0,
) -> FreeFieldCheck

Is this room anechoic or hemi-anechoic enough for ISO 3745? (Annex A as amended).

Judges the inverse-square-law analysis of one or more test sources (inverse_square_law_deviations) against the amended ISO 3745:2012 Annex A and the ISO 26101:2017 clauses it defers to (see the module notes). The radius of A.2.4 is the smallest qualified distance over every traverse and every evaluated frequency; the per-band requirements are judged within it. When the full range fails, the widest contiguous range that meets every requirement is reported as the reduced range A.2.3 allows, “in conformity” but not “in full conformity”.

Parameters

Name	Description
`results`	One `InverseSquareLawResult` per test source, their frequencies disjoint, all for the same room.
`bandwidth`	`"discrete-frequency"` (the default of A.4.1, tones or a narrow-band analysis) or `"broadband"` (noise in one-third octave bands, sufficient only for sources that radiate broadband noise).
`source_directionality`	The `verify_source_directionality` result of each test source (one, or a sequence aligned with `results`); `None` leaves ISO 26101 Annex B unjudged.
`measurement_radius_m`	The radius of the measurement surface to be used, in metres; the room must be qualified at least that far.
`reflecting_plane_absorption_coefficient`	The largest sound absorption coefficient of the reflecting plane over the frequency range (A.2.5), hemi-anechoic only.
`reflecting_plane_margin_m`	How far the reflecting plane extends beyond the projection of the measurement surface, in metres (A.2.5), hemi-anechoic only.
`paths_in_working_area`	Whether the traverse paths lie in the working area of the room, the part normally used for measurements (A.3.3); a declaration, since the positions cannot tell. `None` leaves it unjudged. The targets of A.3.3 a) to e) are read from `MicrophoneTraverse.targets`.
`speed_of_sound`	Speed of sound, in m/s, for the wavelengths of A.4.3, ISO 26101 5.1.4.3 and A.2.5 (default 343).

Returns: FreeFieldCheck.

Raises

Exception	When
ValueError	for results of different rooms, a frequency given to two sources, directionality results that do not align with `results`, an unknown bandwidth, an absorption coefficient outside 0 to 1, a margin that is not finite, or a working-area declaration that is not `True`, `False` or `None`.

directionality_positions

directionality_positions(
    room: FreeFieldRoom,
    *,
    radius_m: float = 1.5,
) -> np.ndarray

Microphone positions of the directionality test (ISO 26101 B.3.2).

The angle of elevation $φ$ is counted from the vertical above the source (90 deg is the reflecting plane), the azimuth $Θ$ from the x axis. The rows run $φ$ = 80, 60, 40 and 20 deg (then 100, 120, 140 and 160 deg in an anechoic room), and within each $Θ$ = 0 to 315 deg in steps of 45 deg: 32 positions in a hemi-anechoic room, 64 in an anechoic one (Figure B.1).

Parameters

Name	Description
`room`	`"anechoic"` or `"hemi-anechoic"`.
`radius_m`	The radius, in metres; 1,5 m in B.3.2.

Returns: (32, 3) or (64, 3) coordinates, in metres.

directionality_tolerance_db

directionality_tolerance_db(
    frequencies_hz: ArrayLike,
    *,
    room: FreeFieldRoom,
) -> np.ndarray

Allowable deviation of the test source directionality (ISO 26101 Table B.1).

Parameters

Name	Description
`frequencies_hz`	One-third octave mid-band frequencies, in hertz.
`room`	`"anechoic"` or `"hemi-anechoic"`.

Returns: The half-width of the tolerance band, in dB, per frequency.

Raises

Exception	When
ValueError	for an unknown room, a non-positive frequency or frequencies that are not one value or a 1-D array.

FreeFieldCheck

FreeFieldCheck(
    bandwidth: str,
    results: tuple[InverseSquareLawResult, ...],
    source_directionality: tuple[SourceDirectionalityResult | None, ...],
    measurement_radius_m: float | None,
    reflecting_plane_absorption_coefficient: float | None,
    reflecting_plane_margin_m: float | None,
    paths_in_working_area: bool | None,
    speed_of_sound: float,
)

Whether a room qualifies as anechoic or hemi-anechoic for ISO 3745.

The verdict of the amended ISO 3745:2012 Annex A over the frequencies evaluated, with the ISO 26101:2017 clauses it defers to. Every per-band array is aligned with frequencies_hz (ascending) and judged within maximum_qualified_radius_m.

Every verdict, the radius and the reduced range are read from the analyses and the declarations the check holds, against the figures the two standards print, so they are not fields: a check cannot be built to qualify a room its traverses do not qualify. The reading is done once, when the check is built.

Attributes

Name	Description
`bandwidth`	`"discrete-frequency"` or `"broadband"` (A.4.1): a room qualified with broadband noise is qualified only for sources that radiate broadband noise.
`results`	The inverse-square-law analysis of each test source.
`source_directionality`	The `verify_source_directionality` result of each test source, aligned with `results`, `None` for a source whose directionality was not judged.
`measurement_radius_m`	The measurement radius to be used, if given.
`reflecting_plane_absorption_coefficient`	The largest sound absorption coefficient of the reflecting plane (A.2.5), or `None`; always `None` in an anechoic room.
`reflecting_plane_margin_m`	How far the reflecting plane extends beyond the projection of the measurement surface (A.2.5), in metres, or `None`; always `None` in an anechoic room.
`paths_in_working_area`	Whether the traverse paths lie in the working area of the room (A.3.3), as declared, or `None`.
`speed_of_sound`	Speed of sound, in m/s, for the wavelengths of A.4.3, ISO 26101 5.1.4.3 and A.2.5.

FreeFieldCheck.background_met

property

Levels at least 6 dB above the background at every point, per band.

ISO 26101 5.1.2.2 c); None when a traverse came without background.

FreeFieldCheck.band_met

property

Per band, whether every judged per-band requirement is met.

Returns: One boolean per evaluated frequency.

FreeFieldCheck.band_radius_m

property

The distance to which each frequency is qualified on every traverse, in metres.

FreeFieldCheck.conforming_radius_m

property

The radius qualified over conforming_range_hz, in metres, or nan.

FreeFieldCheck.conforming_range_hz

property

The widest contiguous range over which every judged requirement is met.

In the A.2.3 sense, (low, high) in hertz, or None; with not_judged empty it is the reduced range A.2.3 lets a report state “in conformity”.

FreeFieldCheck.directionality_met

property

The test source within Table B.1 in the band.

None without a directionality result for its source.

FreeFieldCheck.discrete_frequency

property

Whether the qualification holds for tonal sources too (A.4.1).

Returns: True for a discrete-frequency qualification.

FreeFieldCheck.equal_spacing_met

property

The points of every traverse within the radius equally spaced at the frequency, per band.

A.4.3 and ISO 26101 5.1.4.3; points added between two of them near a peak deviation are allowed (see the module notes for the tolerance).

FreeFieldCheck.frequencies_hz

property

Every evaluated frequency, ascending, in hertz.

FreeFieldCheck.full_frequency_range

property

Whether every frequency A.2.3 requires from 100 Hz to 10 000 Hz was evaluated.

FreeFieldCheck.iso26101_spacing_met

property

Spacing at most a tenth of a wavelength below 1 kHz and 25 mm above, per band.

ISO 26101 A.4.3, cited by A.2.4; the band that contains 1 kHz is held to the stricter of the two.

FreeFieldCheck.length_met

property

Traverse starting at most, and running at least, a quarter wavelength at the lowest frequency, per band.

ISO 26101 5.1.4.3.

FreeFieldCheck.maximum_qualified_radius_m

property

The A.2.4 radius over every evaluated frequency, in metres.

FreeFieldCheck.not_judged

property

The requirements without data, by name.

FreeFieldCheck.passes

property

Whether the room is qualified in full conformity with ISO 3745.

Every requirement judged and met at every evaluated frequency, the whole of 100 Hz to 10 000 Hz evaluated (A.2.3), and the measurement radius, if given, within the qualified one.

Returns: True for a room in full conformity.

FreeFieldCheck.path_angles_met

property

In a hemi-anechoic room, every traverse within 20 deg to 80 deg from the vertical.

The directionality test’s angles (A.3.3); None in an anechoic room, where A.3.3 sets no such limit.

FreeFieldCheck.path_targets_met

property

Every source has traverses towards each of the five targets of A.3.3 a) to e).

None when a source names no target on any traverse.

FreeFieldCheck.plot()

FreeFieldCheck.plot(
    ax: Axes | None = None,
    *,
    language: str = 'en',
    **kwargs: Any,
) -> Axes

Draw the qualified distance per frequency against the radius judged.

Parameters

Name	Description
`ax`	Existing axes, or `None` to create a figure.
`language`	Label language, `"en"` (default) or `"es"`.
`kwargs`	Forwarded to the radius bars.

Returns: The axes.

FreeFieldCheck.points_met

property

At least 10 points on each traverse and 50 in total within the radius (A.4.3), per band.

FreeFieldCheck.reflecting_plane_met

property

A.2.5 in a hemi-anechoic room, None when not judged or in an anechoic room.

FreeFieldCheck.room

property

"anechoic" or "hemi-anechoic", the room of the analyses.

FreeFieldCheck.spacing_met

property

Spacing at most a tenth of a wavelength below 250 Hz and 100 mm above, per band.

Amended ISO 3745 A.4.3; the band that contains 250 Hz is held to the stricter of the two.

FreeFieldCheck.traverse_count_met

property

Five to eight traverses for every source (A.3.3).

FreeFieldCheck.working_area_met

property

The traverse paths lie in the working area of the room (A.3.3), as declared.

None when not declared.

inverse_square_law_deviations

inverse_square_law_deviations(
    traverses: Sequence[MicrophoneTraverse],
    *,
    frequencies_hz: ArrayLike,
    room: FreeFieldRoom,
    origin_m: ArrayLike | None = None,
    source_box_m: ArrayLike | None = None,
) -> InverseSquareLawResult

Deviations from the inverse square law along the traverses of one source.

Applies Formula (1) with the monitor microphone, then Formulae (2) and (4) with the source strength $b$ that maximizes the qualified distance under the Table A.1 limits of room (see the module notes: the midpoint of the admissible interval, an exact rather than an iterative answer), and reports the starting value of Formula (3) beside it.

One call covers one test source: ISO 26101 5.1.2.2 allows two or more to span the frequency range, each with its own mathematical origin, and check_free_field takes one result per source.

Parameters

Name	Description
`traverses`	The straight traverses, all measured at `frequencies_hz`.
`frequencies_hz`	The test frequencies (tones or one-third octave mid-bands), in hertz, one per level column.
`room`	`"anechoic"` or `"hemi-anechoic"`, which selects the Table A.1 limits.
`origin_m`	The mathematical origin of every traverse, `(x, y, z)` in metres; `None` with no `source_box_m` takes the room frame’s origin.
`source_box_m`	The box the test source physically occupies, as `((x_min, y_min, z_min), (x_max, y_max, z_max))` in metres; the origin is then searched inside it (ISO 26101 5.1.3.2, ISO 3745 A.3.3). A zero extent on an axis holds the origin on that coordinate.

Returns: InverseSquareLawResult.

Raises

Exception	When
ValueError	if both `origin_m` and `source_box_m` are given, a traverse’s level columns do not match the frequencies, a traverse has no level at a frequency, or a point coincides with the origin.

inverse_square_law_tolerance_db

inverse_square_law_tolerance_db(
    frequencies_hz: ArrayLike,
    *,
    room: FreeFieldRoom,
) -> np.ndarray

Allowable deviation from the inverse square law per band (Table A.1).

ISO 26101:2017 Annex A and ISO 3745:2012/Amd.1:2017 A.2.2 print the same table. A frequency is read in the one-third octave band that contains it, so a tone at 794 Hz is held to the 800 Hz row.

Parameters

Name	Description
`frequencies_hz`	Test frequencies or band mid-frequencies, in hertz.
`room`	`"anechoic"` or `"hemi-anechoic"`.

Returns: The half-width of the tolerance band, in dB, per frequency.

Raises

Exception	When
ValueError	for an unknown room, a non-positive frequency or frequencies that are not one value or a 1-D array.

InverseSquareLawResult

InverseSquareLawResult(
    frequencies_hz: np.ndarray,
    room: str,
    origin_m: np.ndarray,
    origin_fitted: bool,
    source_box_m: np.ndarray | None,
    traverse_names: tuple[str, ...],
    traverse_targets: tuple[tuple[str, ...], ...],
    positions_m: tuple[np.ndarray, ...],
    distances_m: tuple[np.ndarray, ...],
    levels_db: tuple[np.ndarray, ...],
    background_margin_db: tuple[np.ndarray | None, ...],
    source_stability_db: tuple[np.ndarray, ...],
    source_strength_db: np.ndarray,
    initial_source_strength_db: np.ndarray,
    deviations_db: tuple[np.ndarray, ...],
    traverse_radius_m: np.ndarray,
    band_radius_m: np.ndarray,
)

Deviations from the inverse square law of one test source (ISO 26101 5.1.5).

Attributes

Name	Description
`frequencies_hz`	The test frequencies, in hertz, as given.
`room`	`"anechoic"` or `"hemi-anechoic"`.
`origin_m`	The mathematical origin of the traverses, `(x, y, z)` in metres.
`origin_fitted`	`True` when the origin was searched inside `source_box_m`.
`source_box_m`	The box the test source occupies, `(2, 3)` rows of lower and upper corner, or `None` for a fixed origin.
`traverse_names`	The label of each traverse as given, `""` when none was.
`traverse_targets`	What each traverse was selected towards (A.3.3 a) to e)), as given.
`positions_m`	The measurement points of each traverse, `(N, 3)`.
`distances_m`	The distance $r_{i}$ of each point from the origin, in metres.
`levels_db`	$L_{p i}$ of each traverse after the stability correction of Formula (1), `(N, NF)`, `nan` where not measured.
`background_margin_db`	$L_{p i}^{'}$ less the background at each point, `(N, NF)`, or `None` for a traverse without background.
`source_stability_db`	The largest excursion of the monitor microphone from its reading at point 0, per frequency, or `nan` without a monitor.
`source_strength_db`	$b$ of Formula (2) per traverse and frequency, `(NT, NF)`: the midpoint of the admissible interval over the points within `band_radius_m` (over the traverse’s own qualified run when its first point lies beyond that radius).
`initial_source_strength_db`	The starting value of Formula (3), the mean of $L_{p i} + 20 l g (r_{i} / r_{0})$ over every point, `(NT, NF)`.
`deviations_db`	$Δ L_{p i}$ of Formula (4) at every point, `(N, NF)` per traverse.
`traverse_radius_m`	The qualified distance of each traverse at each frequency, `(NT, NF)`, in metres.
`band_radius_m`	The distance to which each frequency is qualified on every traverse at once, `(NF,)`, in metres.

InverseSquareLawResult.largest_deviation_db

property

The largest $∣Δ L_{p i} ∣$ within each band’s radius.

Returns: One value per frequency, in dB.

InverseSquareLawResult.maximum_qualified_radius_m

property

The largest radius every traverse meets at every frequency (A.2.4).

Returns: The smallest of band_radius_m, in metres.

InverseSquareLawResult.plot()

InverseSquareLawResult.plot(
    ax: Axes | None = None,
    *,
    frequency_hz: float | None = None,
    language: str = 'en',
    **kwargs: Any,
) -> Axes

Draw the deviations along every traverse at one frequency.

Parameters

Name	Description
`ax`	Existing axes, or `None` to create a figure.
`frequency_hz`	The test frequency to draw; `None` draws the one qualified to the shortest distance.
`language`	Label language, `"en"` (default) or `"es"`.
`kwargs`	Forwarded to the deviation curves.

Returns: The axes.

InverseSquareLawResult.tolerance_db

property

The Table A.1 limit of each frequency, in dB, read from the room.

Returns: inverse_square_law_tolerance_db of frequencies_hz in room.

MicrophoneTraverse

MicrophoneTraverse(
    positions_m: np.ndarray,
    levels_db: np.ndarray,
    monitor_levels_db: np.ndarray | None = None,
    background_levels_db: np.ndarray | None = None,
    name: str = '',
    targets: tuple[str, ...] = (),
)

One straight microphone traverse of ISO 26101:2017 5.1.3.2.

Attributes

Name	Description
`positions_m`	The measurement points, one row per point, as `(x, y, z)` coordinates in metres in the room frame: `z` points up and, in a hemi-anechoic room, the reflecting plane is `z = 0`. The mathematical origin of the traverse is given to `inverse_square_law_deviations` in the same frame.
`levels_db`	The measured sound pressure level $L_{p i}^{'}$ at each point, one row per point and one column per test frequency, in dB. A `nan` marks a point not measured at that frequency, which is how a spacing that changes with frequency is written (ISO 3745 A.4.3 asks for equally spaced points at each frequency). Every other level is finite.
`monitor_levels_db`	The monitor microphone level $L_{p, ref, i}$ recorded with each point, same shape; the first row is the level for the initial point 0, which every point is corrected to. It is finite wherever a level was measured and in the first row of every frequency measured. `None` applies no stability correction.
`background_levels_db`	The background level at each point, same shape, or one row for the whole traverse, finite wherever a level was measured; `None` leaves the 6 dB requirement of ISO 26101 5.1.2.2 c) unjudged.
`name`	A label for the traverse (the path it follows: a dihedral corner, the nearest wall, a door).
`targets`	What the path is selected towards, among the five of the amended ISO 3745 A.3.3 a) to e): `"dihedral corner"`, `"trihedral corner"`, `"boundary centre"` (the centre of the most uniform boundary surface), `"closest boundary"` and `"unique features"` (a door, a window, a ventilation opening). One path may serve more than one target, as A.3.3 NOTE 1 allows for c) and d); an empty tuple names none.

MicrophoneTraverse.along()

classmethod

MicrophoneTraverse.along(
    direction: ArrayLike,
    distances_m: ArrayLike,
    levels_db: ArrayLike,
    *,
    start_m: ArrayLike = (0.0, 0.0, 0.0),
    monitor_levels_db: ArrayLike | None = None,
    background_levels_db: ArrayLike | None = None,
    name: str = '',
    targets: Sequence[str] = (),
) -> MicrophoneTraverse

A traverse laid out along a straight line from start_m.

Parameters

Name	Description
`direction`	The direction of the path, any non-zero 3-vector.
`distances_m`	The distance of each point from `start_m` along the path, in metres.
`levels_db`	One row per point, one column per frequency, in dB.
`start_m`	The point the distances are counted from, in metres; the room frame’s origin by default.
`monitor_levels_db`	As for the class.
`background_levels_db`	As for the class.
`name`	As for the class.
`targets`	As for the class.

Returns: The traverse.

Raises

Exception	When
ValueError	for a direction that is zero or not finite.

qualification_frequencies_hz

qualification_frequencies_hz(
    low_hz: float = 100.0,
    high_hz: float = 10000.0,
) -> np.ndarray

The frequencies ISO 3745:2012/Amd.1:2017 A.2.3 evaluates over a range.

Below 125 Hz and above 4 000 Hz every one-third octave band, between them the octave mid-band frequencies 125, 250, 500, 1 000, 2 000 and 4 000 Hz. The default range is the least A.2.3 allows, 100 Hz to 10 000 Hz: eleven frequencies.

Parameters

Name	Description
`low_hz`	Lowest band of the range, in hertz (read in its band).
`high_hz`	Highest band of the range, in hertz.

Returns: The nominal frequencies to evaluate, ascending, in hertz.

Raises

Exception	When
ValueError	if the range is empty.

SourceDirectionalityResult

SourceDirectionalityResult(
    frequencies_hz: np.ndarray,
    room: str,
    mean_level_db: np.ndarray,
    maximum_positive_deviation_db: np.ndarray,
    maximum_negative_deviation_db: np.ndarray,
)

Whether a test source is uniform enough to qualify a room (ISO 26101 B.4).

Attributes

Name	Description
`frequencies_hz`	The one-third octave mid-band frequencies, in hertz.
`room`	The room type the source is to qualify.
`mean_level_db`	The arithmetic mean of the decibel levels per band.
`maximum_positive_deviation_db`	The largest level above the mean.
`maximum_negative_deviation_db`	The largest level below the mean, a negative number.

SourceDirectionalityResult.passes

property

Whether the directionality is within Table B.1 in every band.

Returns: True when every band is within its limit.

SourceDirectionalityResult.plot()

SourceDirectionalityResult.plot(
    ax: Axes | None = None,
    *,
    language: str = 'en',
    **kwargs: Any,
) -> Axes

Draw the extreme deviations per band against Table B.1.

Parameters

Name	Description
`ax`	Existing axes, or `None` to create a figure.
`language`	Label language, `"en"` (default) or `"es"`.
`kwargs`	Forwarded to the deviation bars.

Returns: The axes.

SourceDirectionalityResult.tolerance_db

property

The Table B.1 limit per band, in dB, read from the room.

Returns: directionality_tolerance_db of frequencies_hz in room.

SourceDirectionalityResult.within_tolerance

property

Per band, whether both extreme deviations are within Table B.1.

Returns: One boolean per band.

verify_source_directionality

verify_source_directionality(
    levels_db: ArrayLike,
    *,
    frequencies_hz: ArrayLike,
    room: FreeFieldRoom,
) -> SourceDirectionalityResult

May this source qualify an anechoic or hemi-anechoic room? (ISO 26101 Annex B).

The one-third octave band levels are measured at the 32 positions of a hemisphere or the 64 of a sphere of radius 1,5 m (directionality_positions). Per band, the arithmetic mean of the decibel levels and the largest positive and negative deviations from it are computed (B.3.2); the source is suitable when every deviation is within Table B.1 (B.4). The amended ISO 3745 A.3.1 lets the measurement be made in the room being qualified.

Parameters

Name	Description
`levels_db`	One row per position (32 hemi-anechoic, 64 anechoic), one column per band, in dB; a 1-D array is one band.
`frequencies_hz`	The one-third octave mid-band frequencies, in hertz.
`room`	`"anechoic"` or `"hemi-anechoic"`.

Returns: SourceDirectionalityResult.

Raises

Exception	When
ValueError	for the wrong number of positions, levels that are not finite, or bands that do not match.

Creado y mantenido por José Manuel Requena Plens ORCID Todos los proyectos