noncoaxial op• 数据种类:table → table
• 调用: import fullseye as fs; fs.ledger.ghost_analysis(system, coatings=None, field=None, rings=6, pupil_fill=0.98, pixel_mm=0.005, max_rows=None)(要直接调用实现,import raytrace3d; raytrace3d.ghost_analysis(system, coatings=None, field=None, rings=6, pupil_fill=0.98, pixel_mm=0.005, max_rows=None);从台账取用则 opsoptics.get("ghost_analysis"))
所有面对的两次反射鬼像(table):焦点、在传感器上的大小与相对强度。
> 以下的详细说明为原文 —— 摘要与标题已翻译。
For surface records i < j (refracting surfaces and beamsplitters; nominal mirrors only as the parts of the path they already are), the light runs to *j*, reflects, crosses *j−1 … i+1* backwards, reflects at *i* and runs on to the image plane — N(N−1)/2 paths. The nominal bundle (same launch as the image) is traced along each path with real rays. Per ghost:
• energy_ratio — ghost energy / image energy = R_i·R_j·Π T over the ghost path ÷ Π T over the image path; R from surface_reflectance (the media of the prescription and *coatings*:
a coating for all surfaces, or {record: coating}; records may carry their own coating), at the
angle of incidence of the field's chief ray on each surface (unpolarised; a 45° beamsplitter plate is not
treated as normal incidence);
• focus_offset_mm — the ghost's best focus relative to the image plane (along the image normal; negative = in front of the sensor), radius_mm (RMS) / max_radius_mm on the image plane, centroid_xy (image-plane local) and offset_xy (from the image's chief ray — where the false copy lands relative to the true image);
• point_ratio — ghost irradiance / image irradiance for a point source whose image fills one pixel_mm disk: energy_ratio · (pixel / max(pixel, radius))²;
• extended_ratio — the same for a uniform extended scene: energy_ratio · |det J_image| / |det J_ghost| (the ghost spreads the scene over its own magnification); jacobian is J_ghost (image-plane position of the ghost's chief ray per unit field, 2×2; image_jacobian the image's own).
Rows are sorted by point_ratio (ties by the declared order). n_pairs = N(N−1)/2.
optics 的每个算子都先校验输入再计算(不让任何东西无声通过):
• 单位写进参数名 —— _mm / _um / _deg / _mrad。把 mm 和 µm 弄混不会崩溃,而是给出「看着合理却是错的答案」,所以用命名来防。这里绝不从数值大小去猜单位。
• 字符串一律 ValueError —— float('50') 会成功,于是未解析的配置值会被当成长度混进来(实测:thin_lens('50', '200') 曾返回看着合理的 66.667 mm)。bool 也按 True == 1 的隐式提升拒绝。
• complex / masked array 一律 ValueError(仅接受实数槽位;拒绝无声丢弃虚部或剥掉掩码)。所有输入中的 NaN/Inf 一律 ValueError。
• 逐项点名拒绝除零及其近亲:焦距 0、曲率半径 0、折射率 <= 0、全不透明光阑(全为 0,归一化变成 0/0)、总和 <= 0 的 PSF、S0 = 0 的 Stokes 矢量、物体位于前焦点(像在无穷远)。
• 只有两个算子会返回非有限值,而且都写进了契约:depth_of_field 在超焦距以外返回 far_mm = inf(这正是超焦距的定义),gaussian_beam 在束腰处返回 wavefront_radius_mm = inf(平面波前的曲率半径)。两者都同时返回一个有限的搭档(far_is_infinite / curvature_per_mm)。除此之外的无声 NaN/Inf 都在内部检出并 ValueError ——「float64 溢出了」和「答案是无穷大」是两种不同的主张,不能拿后者的脸去交付前者。
• 尺寸上限:生成网格受 optics.MAX_GRID(4096)限制,传入的场/PSF/光阑受 optics.MAX_FIELD_ELEMENTS(2^24),ABCD 元件序列受 optics.MAX_SYSTEM_ELEMENTS(1024),Zernike 受 MAX_ZERNIKE_TERMS(512)/ MAX_ZERNIKE_ORDER(40)/ MAX_ZERNIKE_BASIS(2^25)。以 fail-closed 堵住「小参数引发巨大内部分配」的路径(实测:n_max=40 × 4096² 需要 108 GB)。
• 物理上不可能的状态同样拒绝:偏振度 > 1 的 Stokes 矢量、负透过率、负强度、n-|m| 为奇数等非法 Zernike 指标。
• 示例数据目录(下载 URL / 许可证) —— 2-D 用 skimage.data(BSD/公有领域)加合成图,3-D 给出真实数据源(Stanford/PDS 等)的下载 URL。
• 算子来历与参考文献 —— 该算子族所依据的研究/方法出处。
• 算法的正典(作者、年份)与用途见上面的族使用指南。
• poc_noncoaxial_optics — py -3.11 examples/poc_noncoaxial_optics.py
table 作为输入)abcd_matrix · wavefront_stats · sfr_from_edge · paraxial_trace · seidel_coefficients · spot_stats · tolerance_analysis · wavefront_from_opd
noncoaxial)noncoaxial_system · noncoaxial_example · noncoaxial_layout · noncoaxial_trace · noncoaxial_spot · image_orientation · beamsplitter_branches · tilted_plate_shift
*Provenance: raytrace3d.py — OPTICS 算子登记表。本条目由 tools/opdocs.py md 自动生成(请勿手工编辑)。*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.