hearing.sound_field_audiometry

The sound field of an audiometric test room (ISO 8253-2:2009).

Sound field audiometry presents the test signal through one or more loudspeakers instead of earphones, so the room becomes part of the instrument. ISO 8253-2 specifies three sound fields a room may be qualified as (Clause 5), each measured with the test subject and the chair absent and with the same signals the audiometry will use (5.1):

• a free sound field (5.2): the loudspeaker at least 1 m from the reference point, the level 0,15 m to the left, right, above and below it within ±1 dB of the level at it up to and including 4 kHz and within ±2 dB above, the two ear-side positions within 3 dB of each other above 4 kHz, and the difference between the points 0,15 m in front of and behind it on the reference axis within ±1 dB of what the inverse distance law gives,

  $$\Delta L_{\mathrm{theory}}=20\lg \frac{r+d}{r-d}\mathrm{dB}$$

  with $r$ the distance from the loudspeaker to the reference point and $d$ = 0,15 m. Its NOTE says only an anechoic room meets it.

• a quasi-free sound field (5.4): the same arrangement with ±2 dB at the four lateral positions and the axial pair at $d$ = 0,10 m. Its usable frequency range is the range in which the requirements are met.

• a diffuse sound field (5.3): the level at six positions 0,15 m from the reference point, on the front-back, right-left and up-down axes, within ±2,5 dB of the level at it and the two ear-side positions within 3 dB of each other; from 500 Hz up, the levels a directional microphone reads at the reference point in the directions of the largest and the smallest incident energy stay within the variation Table 1 allows for the microphone’s front-to-random sensitivity index (5 dB for an index of 5 dB or more, 4,5 dB for 4,5 dB, 4 dB for 4 dB, and a microphone below 4 dB is not suitable).

check_free_sound_field, check_quasi_free_sound_field and check_diffuse_sound_field judge those conditions band by band. The diffuse-field judgement is the same one ISO 4869-3:2007, 5.2.2, asks of the random-incidence field of its test site, with its own Table 1; phonometry.hearing.check_random_incidence_field runs it with that table and returns the same DiffuseSoundFieldCheck.

The ambient noise (Clause 6, Table 2). The maximum permissible ambient sound pressure levels for sound field audiometry, in one-third-octave bands from 31,5 Hz to 12,5 kHz, for a lowest test frequency of 125 Hz or 250 Hz, are SOUND_FIELD_AMBIENT_LIMITS_DB. Its footnote a says they are derived from ISO 8253-1 for binaural listening, and prints no offset; laid side by side, from 31,5 Hz to 8 kHz each cell is the bone-conduction limit of ISO 8253-1:2010 Table 4 less 3 dB, an observation the library makes on the two printed tables and not a figure the footnote gives. Footnote b asks for lower limits, by an amount it does not print, when narrow-band noise is the test signal. phonometry.hearing.check_audiometric_ambient_noise judges a measured spectrum against them with presentation="sound field".

Off-axis loudspeakers (Annex B, informative). No standardized reference threshold exists for a loudspeaker away from 0° incidence (4.8, NOTE 1). Table B.1 prints how much higher the sound pressure level is at the ear closest to the loudspeaker at 45° and 90°, rounded to the nearest 0,5 dB, which INCIDENCE_CORRECTIONS_DB carries and incidence_correction reads. The paragraph above the table announces it from 200 Hz; its rows start at 125 Hz, and the library carries every printed row (see docs/ERRATA.md).

Clause, table and annex numbers refer to ISO 8253-2:2009(E).

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

check_diffuse_sound_field(
    position_levels_db: Mapping[str, ArrayLike],
    reference_levels_db: ArrayLike,
    *,
    directional_levels_db: ArrayLike | None = None,
    front_to_random_index_db: ArrayLike | None = None,
    frequencies: ArrayLike | None = None,
) -> DiffuseSoundFieldCheck

Is the sound field of the test room diffuse? (ISO 8253-2, 5.3).

Measured with the test subject and the chair absent, with the signals the audiometry will use:

• a) the level at each of the six positions 0,15 m from the reference point, front and back, left and right, up and down, read with an omnidirectional microphone kept in one orientation, stays within ±2,5 dB of the level at the reference point, and the extreme right and left positions within 3 dB of each other;
• b) from 500 Hz up, the levels a directional microphone reads at the reference point in the directions of the largest and smallest incident energy stay within the variation Table 1 allows for its front-to-random sensitivity index (DIFFUSE_FIELD_VARIATION_LIMITS).

The readings of b) are taken in as many directions as the microphone and the loudspeakers need, including the two planes where the extremes are expected (footnote a); the variation is the largest less the smallest of them.

Parameters

Returns: DiffuseSoundFieldCheck.

Raises

check_free_sound_field(
    lateral_levels_db: Mapping[str, ArrayLike],
    reference_levels_db: ArrayLike,
    *,
    front_levels_db: ArrayLike,
    back_levels_db: ArrayLike,
    loudspeaker_distance_m: float,
    frequencies: ArrayLike | None = None,
) -> FreeSoundFieldCheck

Is the sound field of the test room free? (ISO 8253-2, 5.2).

Measured with the test subject and the chair absent, the loudspeaker at the head height of a seated listener with its reference axis through the reference point:

• a) the loudspeaker is at least 1 m from the reference point;

• b) the level 0,15 m from the reference point to the left, right, above and below it stays within ±1 dB of the level at it up to and including 4 kHz and within ±2 dB above, and above 4 kHz the right and left positions differ by 3 dB at most;

• c) the level at the point 0,15 m in front of the reference point on the reference axis less the level 0,15 m behind it stays within ±1 dB of the inverse distance law,

  $$\Delta L_{\mathrm{theory}}=20\lg \frac{r+0,15\mathrm{m}}{r-0,15\mathrm{m}}\mathrm{dB}$$

  with $r$ the distance from the loudspeaker to the reference point, the point source the law assumes.

Pure tones may only be used in a field that meets this (4.2), and the NOTE to 5.2 says only an anechoic room does.

Parameters

Returns: FreeSoundFieldCheck.

Raises

check_quasi_free_sound_field(
    lateral_levels_db: Mapping[str, ArrayLike],
    reference_levels_db: ArrayLike,
    *,
    front_levels_db: ArrayLike,
    back_levels_db: ArrayLike,
    loudspeaker_distance_m: float,
    frequencies: ArrayLike | None = None,
) -> FreeSoundFieldCheck

Is the sound field of the test room quasi-free? (ISO 8253-2, 5.4).

The field a room that is not anechoic can still offer, measured with the test subject and the chair absent and all other working conditions kept:

• a) the loudspeaker’s reference point is at least 1 m from the reference point;

• b) the level 0,15 m from the reference point to the left, right, above and below it stays within ±2 dB of the level at it;

• c) the level 0,10 m in front of the reference point on the reference axis less the level 0,10 m behind it stays within ±1 dB of the inverse distance law,

  $$\Delta L_{\mathrm{theory}}=20\lg \frac{r+0,10\mathrm{m}}{r-0,10\mathrm{m}}\mathrm{dB}$$

The usable frequency range of the field is the range in which these hold, which FreeSoundFieldCheck.usable_frequencies gives.

Parameters

Returns: FreeSoundFieldCheck with field="quasi-free".

Raises

Constant (tuple).

DIFFUSE_FIELD_VARIATION_LIMITS = ((5.0, 5.0), (4.5, 4.5), (4.0, 4.0))

DiffuseSoundFieldCheck(
    frequencies: np.ndarray,
    position_deviation_db: np.ndarray,
    left_right_difference_db: np.ndarray,
    directional_variation_db: np.ndarray,
    front_to_random_index_db: np.ndarray | None,
    standard: str,
    positions: tuple[str, ...] = ('front', 'back', 'left', 'right', 'up', 'down'),
)

Whether a sound field is diffuse enough, band by band.

The verdict of ISO 8253-2:2009, 5.3, and of the random-incidence field of ISO 4869-3:2007, 5.2.2, which ask the same two things and differ only in the Table 1 that turns a microphone’s front-to-random sensitivity index into the variation it may read.

Attributes

What Table 1 allows the variation (allowable_variation_db) is read from the microphone’s index and the standard’s table, not stored, so a check cannot be built against another table.

The directional test is a requirement, not an option: a band from 500 Hz up without a suitable microphone’s reading is not judged, directionality_judged says so, and passes is False until it is. uniform and balanced still give the verdict of the six positions on their own.

property

What the Table 1 allows the directional variation per band, in dB.

Returns: The allowance, or NaN where it is not judged: below 500 Hz, without a directional reading, or where the microphone’s index is below the table’s last row.

property

Per band, whether right and left differ by 3 dB at most.

Returns: One boolean per band.

property

Per band, whether the directional variation stays within Table 1.

Bands the test does not reach, below 500 Hz, count as meeting it; a band from 500 Hz up that was not judged reads False.

Returns: One boolean per band.

property

Per band, whether the directional test was judged.

Returns: True where a suitable microphone’s reading was given, and below 500 Hz, where the test does not apply.

property

Per band, whether the directional test applies (500 Hz and up).

Returns: One boolean per band.

property

Whether every band from 500 Hz up was judged by a suitable microphone.

Returns: True when none is left unjudged.

property

Whether the field qualifies as diffuse, every requirement judged.

Returns: True when every band meets the six positions, the right-left difference and the directional test.

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

Draw each condition against its limit, band by band.

Parameters

Returns: The axes.

property

Per band, whether all six positions stay within ±2,5 dB.

Returns: One boolean per band.

FreeSoundFieldCheck(
    field: str,
    frequencies: np.ndarray,
    lateral_deviation_db: np.ndarray,
    axis_difference_db: np.ndarray,
    loudspeaker_distance_m: float,
    positions: tuple[str, ...] = ('left', 'right', 'up', 'down'),
)

Whether a free or quasi-free sound field is met, band by band.

The verdict of ISO 8253-2:2009, 5.2 (free) or 5.4 (quasi-free).

Attributes

Where the axial points sit (axis_offset_m) is the clause’s, and so is the difference the inverse distance law gives there (inverse_distance_difference_db): both are read from the field and the loudspeaker distance and are not fields, so a check cannot be built against another offset or another law.

property

$d$, how far the axial points sit from the reference point.

Returns: 0,15 m for a free field (5.2 c)), 0,10 m for a quasi-free one (5.4 c)), in metres.

property

Per band, whether right and left differ by 3 dB at most above 4 kHz.

5.2 b) asks it of a free field above 4 kHz only; below, and for a quasi-free field, the band reads True.

Returns: One boolean per band.

property

Per band, whether every positional requirement is met.

Returns: One boolean per band.

property

Whether the loudspeaker is at least 1 m from the reference point.

Returns: True when 5.2 a) or 5.4 a) is met.

property

Per band, whether the axial difference is within ±1 dB of the law.

Returns: One boolean per band.

property

The axial difference less what the inverse distance law gives, in dB.

Returns: One value per band.

property

What the inverse distance law gives for the axial difference, in dB.

Returns: $20\lg ((r+d)/(r-d))$ with $r$ the loudspeaker distance and $d$ the axial offset.

property

The tolerance of the four lateral positions per band, in dB.

Returns: ±1 dB up to and including 4 kHz and ±2 dB above for a free field (5.2 b)), ±2 dB throughout for a quasi-free one (5.4 b)).

property

The difference between the right and left positions, in dB.

Returns: Its absolute value per band.

property

Whether the field qualifies at every test frequency given.

Returns: True when the loudspeaker distance and every band meet the clause.

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

Draw the lateral and axial conditions against their limits.

Parameters

Returns: The axes.

property

Per band, whether all four lateral positions stay within tolerance.

Returns: One boolean per band.

property

The frequencies at which the field meets its requirements, in hertz.

For a quasi-free field this is its usable frequency range (5.4).

Returns: The frequencies of the compliant bands.

incidence_correction(
    frequencies: ArrayLike,
    *,
    incidence_angle_deg: float,
) -> np.ndarray

How much louder the nearer ear is with the loudspeaker off axis (Table B.1).

The increase in sound pressure level at the ear closest to the loudspeaker at 45° or 90° incidence, relative to 0°, rounded to the nearest 0,5 dB as Annex B prints it. The annex is informative and says no standardized reference threshold exists for these angles (4.8, NOTE 1); how a report uses the increase is left to it.

Parameters

Returns: The increase per frequency, in dB.

Raises

Constant (tuple).

INCIDENCE_CORRECTION_FREQUENCIES_HZ = (125.0, 160.0, 200.0, 250.0, 315.0, 400.0, 500.0, 630.0, 800.0, 1000.0, 1250.0, 1500.0, 1600.0, 2000.0, 2500.0, 3000.0, 3150.0, 4000.0, 5000.0, 6000.0, 6300.0, 8000.0, 10000.0, 12500.0)

Constant (mapping).

INCIDENCE_CORRECTIONS_DB = {45.0: array([0.5, 1. , 1. , 1. , 1.5, 2.5, 3. , 3.5, 3.5, 4. , 4. , 3.5, 3.5,
       3. , 3.5, 5. , 5. , 4. , 6. , 7.5, 7.5, 5.5, 4.5, 1.5]), 90.0: array([ 1. ,  1.5,  1.5,  2. ,  2.5,  3.5,  4.5,  5. ,  5. ,  5.5,  6. ,
        5. ,  4.5,  2. ,  2. ,  2.5,  2. , -0.5,  4. ,  9.5, 10. ,  8.5,
        6. ,  8. ])}

Constant (tuple).

SOUND_FIELD_AMBIENT_BANDS_HZ = (31.5, 40.0, 50.0, 63.0, 80.0, 100.0, 125.0, 160.0, 200.0, 250.0, 315.0, 400.0, 500.0, 630.0, 800.0, 1000.0, 1250.0, 1600.0, 2000.0, 2500.0, 3150.0, 4000.0, 5000.0, 6300.0, 8000.0, 10000.0, 12500.0)

Constant (mapping).

SOUND_FIELD_AMBIENT_LIMITS_DB = {125.0: array([52., 44., 38., 32., 27., 22., 17., 14., 12., 10.,  8.,  6.,  5.,
        5.,  4.,  4.,  4.,  5.,  5.,  3.,  1., -1.,  1.,  6., 12., 14.,
       15.]), 250.0: array([60., 53., 46., 41., 36., 32., 25., 18., 12., 10.,  8.,  6.,  5.,
        5.,  4.,  4.,  4.,  5.,  5.,  3.,  1., -1.,  1.,  6., 12., 14.,
       15.])}