wave opimage2d × image2d → image2dimport fullseye as fs; fs.ledger.pupil_blur(image, pupil, defocus_waves=0.0, wavelength_um=0.55, f_number=5.6, pixel_pitch_um=5.0, oversample=4, opd_waves=None) (実装を直接呼ぶなら import optics; optics.pupil_blur(image, pupil, defocus_waves=0.0, wavelength_um=0.55, f_number=5.6, pixel_pitch_um=5.0, oversample=4, opd_waves=None)、台帳から引くなら opsoptics.get("pupil_blur"))Blur an image with the PSF of a pupil shape at one wavelength band.
The forward imaging model of one spectral band: the PSF of
:func:pupil_psf (same pupil / defocus / wavelength / f-number
arguments), binned to the image’s pixel pitch so the blur is in the
image’s own units, convolved with the image (FFT, reflect-padded borders so
a flat field stays flat and nothing wraps around). Call it once per band
with that band’s defocus_waves (from :func:defocus_from_shift and a
focal shift versus wavelength) and you have the polychromatic image of a
lens with longitudinal chromatic aberration seen through any pupil — the
ingredient of the “colour from chromatic blur” hypothesis for
single-photoreceptor eyes (Stubbs & Stubbs, PNAS 113:8206, 2016).
Returns a float64 image2d of the image’s shape. Linear: no clipping,
no re-normalisation of the image (the PSF sums to 1, so a constant image
is returned unchanged to rounding).
Ground truth it reproduces (measured, tests/test_optics.py): a constant
image is unchanged to 1e-12; a delta image returns the binned PSF itself
(the impulse response, to 1e-12 where the kernel fits); with
defocus_waves = 0 and a diffraction spot much smaller than the pixel
(lambda N = 0.8 um on a 5 um pitch) a sharp edge is unchanged to
within 1 % — the identity at focus; the blurred image’s total is the
input’s total to 1e-9 (flux is conserved by the reflect padding).
Raises ValueError: everything :func:pupil_psf raises, plus
image not 2-D / smaller than 2x2 / over the size cap / complex / masked
/ non-finite, and a non-finite result (an FFT overflow).
Shift-invariant: one PSF for the whole field. Field-dependent blur
(vignetting, off-axis aberration) is lensimage.render_through_lens.
optics の全 op は入力を検証してから計算する(黙って通さない):
_mm / _um / _deg / _mrad。mm と µm の取り違えは crash ではなく「もっともらしく間違った答え」なので、名前で防ぐ。大きさから単位を推測する処理は一切しない。ValueError — float('50') は成功してしまうため、未パースの設定値が長さとして通り抜ける(実測: thin_lens('50', '200') がもっともらしい 66.667 mm を返していた)。bool も True == 1 の暗黙昇格として拒否。ValueError(実数枠のみ。虚部の無言切り捨て・マスク剥がしを拒否)。NaN/Inf は全入力で ValueError。depth_of_field の過焦点距離以遠の far_mm = inf(それが過焦点距離の定義)と gaussian_beam のウエストでの wavefront_radius_mm = inf(平面波面の曲率半径)。どちらも有限の相棒(far_is_infinite / curvature_per_mm)を併せて返す。それ以外の無言 NaN/Inf は内部で検出して ValueError —「float64 が溢れた」と「答えが無限大」は別の主張なので、後者の顔で前者を返さない。optics.MAX_GRID(4096)、供給された場/PSF/開口は optics.MAX_FIELD_ELEMENTS(2^24)、ABCD 素子列は optics.MAX_SYSTEM_ELEMENTS(1024)、Zernike は MAX_ZERNIKE_TERMS(512)/ MAX_ZERNIKE_ORDER(40)/ MAX_ZERNIKE_BASIS(2^25)。小さな引数から巨大な内部確保が起きる経路(実測: n_max=40 × 4096² で 108 GB)を fail-closed で塞ぐ。| 物理的に不可能な状態も拒否: 偏光度 > 1 の Stokes ベクトル、負の透過率、負の強度、n- | m | が奇数などの不正な Zernike 添字。 |
py -3.11 examples/optics_imaging.pyimage2d を入力に取れる)fraunhofer_pattern · pupil_psf · psf_to_mtf · illumination_uniformity · render_through_lens · surface_defect · defocus_blur
wave)airy_pattern · angular_spectrum_propagate · fraunhofer_pattern · gaussian_beam · defocus_from_shift · pupil_psf
Provenance: optics.py — OPTICS operator registry. この per-op ノートは tools/opdocs.py md が自動生成(手編集しない)。
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.