jmrp.io / PyOctaveBand

Fork 0

[READ-ONLY] Mirror of https://github.com/jmrplens/PyOctaveBand. [Python3] Octave-Band and Fractional Octave-Band filter. For signal in time domain. jmrplens.github.io/PyOctaveBand/

acoustics audio filter frequency frequency-analysis frequency-domain octave python3 signal time-domain

Fork 0

PyOctaveBand / tests / data / norah2 /

at main 2 files

José M. Requena Plens Derive the hover and idle sources from the ground hover ring (#539) 2mo ago

02c29d38

README.md

Derive the hover and idle sources from the ground hover ring (#539) * Derive the hover and idle sources from the ground hover ring (NORAH2 Table 3) The guidance's A.3.5 measures in-ground hover on a ring of ground microphones and derives the out-of-ground hover and idle sources from it. hover_ring_hemisphere extends the ring to a hemisphere assuming constant directivity in phi, with the reference implementation's horizontal-bearing lookup as an explicit mapping option; hover_derived_hemisphere applies the Table 3 offsets, the published constants as defaults and any measured or database correction as a parameter. The hemisphere plot now labels its own reference distance instead of assuming 60 m. * Reproduce the prototype's out-of-ground-hover case end to end Extend the committed NORAH2 verification subset with the three files of the ARP Case 4 R22_DEP_RT2 operation (47 files, byte-identical to the public release) and anchor the ring derivation to the real data: the committed hover ring reproduces on the hemisphere rim band for band, and the eight microphones of the OGH case reproduce LASmax and SEL to 0.45 dB worst case with the prototype's bearing mapping and its +8 dB database correction passed explicitly. * Record the A.3.5 errata and the Table 3 conformance claims Three registry entries for the rotorcraft guidance: the Approach 3 prose derives full-rpm idle from out-of-ground hover where its own Table 3 derives it from in-ground hover; the taxi sentence pairs wheeled and wheel-less helicopters with the transposed sources; and the published +12/-12/-2.5 dB offsets disagree with the +8/-10/-2 dB corrections shipped in every interpolation file of the public database. All three verified on the rendered pages. Two conformance claims pin the constant-phi ring derivation and the Approach 3 HOGE offset (535 checks). * Document the hover, idle and taxi derivation A new section of the rotorcraft guide (docs mirror, site English and Spanish) walks the ring measurement, the constant-phi extension, the Table 3 offsets and the taxi selection rule, with a new figure deriving the four sources from a synthetic ring. The scope notes now leave only the no-hover-data side-line fallback outside the implementation. API reference and llms files regenerated. * Polish the hover figure and the Table 3 notes The hover-ring figure now declares its right panel's phi = 0 section, sets the idle offsets with a real minus sign and describes the ring as it is, peaking at the nose and falling to the tail with the starboard half above the port half (generator comment, alt texts and their Spanish versions). The verification notes count the four flow resistivities of the hover case, the snippet comments read as one 31-band spectrum per bearing, and the shipped-corrections erratum names the triangulation lookup files the way the A.3.1 entry does. * Close the review comments of the hover derivation The aircraft overview now counts the no-hover-data side-line fallback as the remaining rotorcraft gap instead of hover, idle and taxi; the guide's hover snippets define a representative ring and the wheeled flag so every name resolves on the page; the errata quotes escape their literal asterisks; the conformance pages carry a single JSON-LD object per language (main shipped it duplicated); the hemisphere plot keeps the precision of a non-integer reference distance; and hover_ring_hemisphere rejects non-positive band frequencies. * Select the phi = 0 hover column by index, not by float equality The azimuth axis of the derived hover hemisphere is symmetric by construction, so each node's side is fixed by its position: port before the middle, starboard after it and, with an even cell count, the phi = 0 column exactly in the middle. Selecting the column by index instead of comparing the node against 0.0 removes the floating-point equality and keeps the energy-mean column in place even for a step whose middle node lands a few ulp away from zero; a boundary test pins a 180/14 degree step.

2 months ago

norah2_arp_extract.zip

2 months ago

README.md

NORAH2 ARP verification subset — rotorcraft oracle data#

What this folder contains#

norah2_arp_extract.zip holds the verification subset of the NORAH2 public release (V2.0.74) consumed by tests/aircraft/test_rotorcraft_norah2.py: the R22 hemisphere set with its triangulation lookup (Hemispheres/R22_*, including the in-ground-hover ring measurement) and the ARP single-event cases used as oracles (input flight paths .inp, per-step microphone histories .his and per-microphone metrics .onl for the Case 2, Case 3 and Case 4 operations listed in that module, among them the Case 4 out-of-ground-hover operation R22_DEP_RT2). The internal layout of the original archive is preserved, so the same parser code reads either source.

It is a byte-identical extract of 47 of the ~360 files of the public release; nothing is modified, resampled or re-encoded.

Source and authorship#

Extracted from the NORAH2 V2.0.74 public release, the prototype and database published by the EASA research project EASA.2020.FC.06 (NORAH, noise of rotorcraft assessed by a hemisphere method). The rotorcraft module of this library is a clean-room implementation of ECAC Doc 32 and the NORAH2 modelling guidance (SC01.D1.5d) and shares no code with the prototype.
The extract carries only measurement-derived model data (source-noise hemispheres and verification-case inputs/outputs), no software: the prototype executables and the full database are not committed.

Purpose and scope of use#

The subset lets the end-to-end oracle suite run everywhere, including CI cells without access to the full archive, validating the flight-condition interpolation, kinematics/retarded time, rotorcraft_event_level and rotorcraft_noise_contour against the published prototype outputs. When the full public release is available locally the suite prefers it: an extraction pointed at by NORAH2_DATA, or tests/data-local/norah2/NORAH2_V2.0.74_public.zip (see tests/data/README.md). Every case reads the same files from either source, so the assertions do not differ. The data are not part of the phonometry package and are not installed with it.

Removal policy#

If you represent EASA or the NORAH project consortium and consider that committing this verification subset exceeds the intended use of the public release, please open an issue or contact the maintainer (see CITATION.cff) and it will be removed promptly, together with the CI use of the tests that read it.

Configure Feed