weighting_response — ACOUSTICS level op

Datenarten: signalsignal

Aufruf: import acoustics; acoustics.weighting_response(freqs, kind='A', floor_db=-200.0) (oder opsacoustics.get("weighting_response"))

Verwendung

Die A-/C-/Z-Frequenzbewertungskurve in dB bei den angegebenen Frequenzen.

> Die ausführliche Beschreibung unten ist der Originaltext — Zusammenfassung und Überschriften sind übersetzt.

Computed from the four pole frequencies that *define* the networks and

normalised so that the response at 1 kHz is exactly 0 dB by construction

— the curve is divided by its own value at 1 kHz rather than having a

published offset constant added to it. That is why the tests can assert

equality at 1 kHz to 0.0 rather than to a tolerance, and why no standard's

table of attenuations appears anywhere in this repository.

The response depends on `f only through f**2`, so it is an even

function and negative frequencies are evaluated at `|f|` — that is the

definition, not a repair. `f = 0` has zero response (both curves have a

zero at DC) and is reported as `floor_db rather than -inf`.

Measured (computed, then printed — these are outputs, not transcriptions):

======== ========= =========

f (Hz) A (dB) C (dB)

======== ========= =========

10 -70.4304 -14.3300

31.5 -39.5250 -3.0305

100 -19.1428 -0.2996

1000 0.0000 0.0000

4000 0.9633 -0.8260

10000 -2.4918 -4.4055

20000 -9.3469 -11.2786

======== ========= =========

`A(1000) and C(1000) are exactly 0.0` — the Python float, not a

rounding — because of the construction. The low-frequency asymptote is a

closed form and is asserted in the tests: `A` falls at exactly

80 dB/decade as `f -> 0 (f**4 over three constants) and C` at

exactly 40 dB/decade (`f**2`). Measured between 0.001 and 0.01 Hz with the

floor lowered out of the way: 79.999998 and 39.999998 dB/decade.

That last caveat is real and is why the floor is an argument: with the

default `floor_db = -200` the A curve reaches the floor below about

0.35 Hz (unfloored, `A(0.1) = -228.55` dB), so the asymptote measured

against the default floor comes out as 0.0 dB/decade between 0.01 and

0.1 Hz — a clamp, correctly reported, that would look like a bug if the

floor were not visible.

Returns a float64 array the same shape as *freqs*.

Raises `ValueError`: a non-1-D / non-finite / complex / masked

`freqs, an unknown kind`.

Ausführlicher Anwendungsleitfaden

Leitfaden zur Familie acoustic_condition_monitoring

Referenzen (Beispieldaten, Literatur)

• Katalog der Beispieldaten (Download-URLs / Lizenzen) — 2-D nutzt skimage.data (BSD/Public Domain) plus synthetische Bilder, 3-D nennt Download-URLs echter Datenquellen (Stanford, PDS, …).

• Herkunft und Literatur der Operatoren — die Quellen der Forschung/Verfahren, auf denen diese Operatorfamilie beruht.

• Der kanonische Algorithmus (Autor, Jahr) und seine Anwendungen stehen im Familienleitfaden oben.

Ausführbare Beispiele (verifizierte Samples, die diesen Operator wirklich aufrufen)

acoustic_condition_monitoringpy -3.11 examples/acoustic_condition_monitoring.py

Typkompatible Folge-Operatoren (nehmen signal als Eingabe)

stft · envelope_spectrum · spectral_kurtosis · cepstrum · angular_resample · order_spectrum · octave_spectrum · apply_weighting

Gleiche Kategorie (level)

octave_bands · octave_spectrum · apply_weighting · equivalent_level · percentile_level


*Provenance: acoustics.py — ACOUSTICS Operator-Registry. Diese Notiz wird von tools/opdocs.py md erzeugt (nicht von Hand bearbeiten).*

© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.