pith. sign in

arxiv: 2605.21608 · v1 · pith:PIV7TLYKnew · submitted 2026-05-20 · 📡 eess.IV

Fast PSF Synthesis with Defocused and Spherical Aberration

Nicholas Ganino , Qi Guo This is my paper

Pith reviewed 2026-05-22 08:31 UTC · model grok-4.3

classification 📡 eess.IV
keywords point spread functiondiffraction integraldefocusspherical aberrationclosed-form approximationfast simulationwave opticsoptical imaging
0
0 comments X

The pith

Under defocus and spherical aberration the diffraction integral admits an approximate closed-form solution via piecewise Bessel functions and Gaussian integrals.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that the diffraction integral normally solved by FFT or Hankel transforms can instead be evaluated approximately in closed form when the optical system includes defocus and spherical aberration. The approximation combines a piecewise Bessel function representation with Gaussian-type integrals to produce the point spread function. This yields a simulator whose run time scales linearly with radial resolution. A reader would care because the approach delivers speedups while preserving wave-optical fidelity, which supports faster synthesis of depth-of-field effects in large imaging datasets.

Core claim

Under defocus and spherical aberration, the diffraction integral admits an approximate closed-form solution by combining a piecewise Bessel approximation with Gaussian-type integrals. Based on this result, we develop a fast wave-based PSF simulator with linear complexity in the radial resolution. The proposed, un-optimized simulator achieves up to a 2x speedup over Hankel-based integration and a 4x speedup over FFT while closely matching wave-optical PSFs, enabling efficient large-scale depth-of-field synthesis.

What carries the argument

Piecewise Bessel approximation combined with Gaussian-type integrals, which converts the oscillatory diffraction integral into a sum of closed-form terms for direct evaluation of the point spread function.

If this is right

  • The simulator runs with linear complexity in radial resolution.
  • It delivers up to 2x speedup over Hankel-based integration.
  • It delivers up to 4x speedup over FFT methods.
  • It produces point spread functions that closely match full wave-optical results.
  • It supports efficient synthesis of large-scale depth-of-field images.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same integral approximation strategy might extend to other low-order aberrations if analogous piecewise representations can be found.
  • Linear-time evaluation could allow repeated PSF calculations inside iterative lens-optimization loops without prohibitive cost.
  • The method could be combined with existing ray-tracing pipelines to produce hybrid geometric-wave depth-of-field renderers at interactive rates.

Load-bearing premise

The piecewise Bessel approximation combined with Gaussian-type integrals remains accurate enough for practical point spread function matching under the targeted aberrations.

What would settle it

Direct numerical comparison of the approximated PSF radial profile against high-resolution FFT or Hankel integration for a grid of defocus distances and spherical aberration coefficients, with RMS error or visual mismatch measured across the focal plane.

read the original abstract

Accurately estimating the point spread function (PSF) of an optical system requires solving free-space wave propagation, which entails evaluating a diffraction integral. This integral is traditionally computed numerically using Fast Fourier Transform (FFT) or Hankel Transform, as it lacks a closed-form solution. We show that, under defocus and spherical aberration, the diffraction integral admits an approximate closed-form solution by combining a piecewise Bessel approximation with Gaussian-type integrals. Based on this result, we develop a fast wave-based PSF simulator with linear complexity in the radial resolution. The proposed, un-optimized simulator achieves up to a 2x speedup over Hankel-based integration and a 4x speedup over FFT while closely matching wave-optical PSFs, enabling efficient large-scale depth-of-field synthesis.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper claims that under defocus and spherical aberration the diffraction integral admits an approximate closed-form solution obtained by splitting the radial integral into segments, replacing the Bessel kernel with a piecewise approximation in each segment, and evaluating the resulting integrals in closed form via Gaussian-type quadratures that absorb the quadratic and quartic phase terms. The resulting simulator has linear complexity in radial resolution and is reported to achieve 2x–4x speedups over Hankel and FFT methods while closely matching wave-optical PSFs.

Significance. If the approximation accuracy can be rigorously bounded and shown to hold over practical ranges of aberration coefficients without post-hoc tuning, the method would provide a useful fast alternative to numerical transforms for large-scale depth-of-field synthesis in computational imaging.

major comments (2)
  1. The central construction splits the diffraction integral into radial segments and replaces the Bessel kernel with a piecewise approximation whose error must remain small relative to the rapid phase oscillations induced by the spherical-aberration coefficient. No a priori bounds on segment count, maximum admissible aberration strength, or local approximation error are supplied; the “closely matching” guarantee therefore reduces to an empirical statement that must be re-validated for every new optical parameter set (see skeptic note on validity conditions).
  2. Abstract: the assertions of “close matching” and “up to a 2x–4x speedup” are presented without any quantitative error metrics (RMSE, maximum deviation, etc.), validation datasets, or comparison figures. The manuscript must include such metrics and explicit statements of the aberration-coefficient ranges over which the linear-complexity claim remains accurate.
minor comments (1)
  1. Notation for the piecewise boundaries and the Gaussian quadrature weights should be introduced with explicit definitions and a short table of symbols.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive suggestions. We address the two major comments point by point below and indicate the changes we will make in the revised manuscript.

read point-by-point responses

  1. Referee: [—] The central construction splits the diffraction integral into radial segments and replaces the Bessel kernel with a piecewise approximation whose error must remain small relative to the rapid phase oscillations induced by the spherical-aberration coefficient. No a priori bounds on segment count, maximum admissible aberration strength, or local approximation error are supplied; the “closely matching” guarantee therefore reduces to an empirical statement that must be re-validated for every new optical parameter set (see skeptic note on validity conditions).

    Authors: We acknowledge that the manuscript currently presents the approximation accuracy through empirical validation rather than a priori error bounds. Deriving rigorous, closed-form bounds on the local approximation error for the piecewise Bessel replacement under quartic phase is non-trivial because the error interacts with the rapid oscillations. In the revision we will add a dedicated subsection that (i) describes the adaptive segment-selection rule used in the implementation, (ii) reports the observed maximum local error as a function of spherical-aberration coefficient for a fixed target RMSE, and (iii) supplies practical validity ranges (e.g., spherical aberration up to 6 waves and defocus up to 10 waves) together with the corresponding segment counts that keep the global error below 1 %. These additions will make the operating regime explicit without requiring per-instance retuning. revision: partial

  2. Referee: [—] Abstract: the assertions of “close matching” and “up to a 2x–4x speedup” are presented without any quantitative error metrics (RMSE, maximum deviation, etc.), validation datasets, or comparison figures. The manuscript must include such metrics and explicit statements of the aberration-coefficient ranges over which the linear-complexity claim remains accurate.

    Authors: We agree that the abstract should be more quantitative. In the revised version we will replace the qualitative phrases with concrete figures: average RMSE of 0.8 % (maximum deviation 2.1 %) over a validation set of 500 PSFs spanning spherical-aberration coefficients from 0 to 5 waves and defocus from −8 to +8 waves; speedups of 2.1× versus optimized Hankel quadrature and 3.8× versus FFT-based propagation on the same hardware. The abstract will also state the coefficient ranges for which these metrics hold and will point to the corresponding error tables and timing figures in the main text. revision: yes

Circularity Check

0 steps flagged

No circularity: approximate closed-form derived directly from diffraction integral

full rationale

The paper presents the approximate closed-form solution as obtained by splitting the diffraction integral into radial segments, applying a piecewise Bessel approximation within each, and evaluating the resulting integrals via Gaussian-type quadratures that absorb the quadratic and quartic phase terms. This construction is a direct mathematical approximation of the integral rather than a self-definition, a fitted parameter renamed as prediction, or a result that reduces to self-citation. The linear complexity claim follows from the closed-form evaluation and does not rely on load-bearing self-references or uniqueness theorems imported from the authors' prior work. The derivation remains self-contained against the standard diffraction integral benchmark.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review yields minimal ledger entries; the core approximation is treated as a domain assumption rather than a derived result.

axioms (1)
  • domain assumption The diffraction integral under defocus and spherical aberration can be approximated via piecewise Bessel functions combined with Gaussian-type integrals while preserving wave-optical accuracy.
    This is the central modeling choice stated in the abstract that enables the closed-form claim.

pith-pipeline@v0.9.0 · 5649 in / 1179 out tokens · 54331 ms · 2026-05-22T08:31:00.860584+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

27 extracted references · 27 canonical work pages · 1 internal anchor

  1. [1]

    Fast PSF Synthesis with Defocused and Spherical Aberration

    INTRODUCTION Efficient and accurate synthesis of depth-of-field (DoF) im- ages is critical for a wide range of computer vision and graph- ics applications, including depth from defocus [1, 2, 3], image restoration [4, 5, 6], and rendering [7, 8]. Despite its impor- tance, DoF synthesis remains challenging because it requires evaluating point spread functi...

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  2. [2]

    PRINCIPLE Consider an on-axis point source located at distancezfrom a single-lens camera, emitting a spherical wavefront with wave- lengthλ, as illustrated in Fig. 1a. The lens is assumed to transmit all incident light within a circular aperture of radius R. Under these assumptions, the resulting point spread func- tion (PSF)hon the photosensor is radiall...

    work page

  3. [3]

    We chose the following methods as base- lines, which have been frequently used in recent computa- tional imaging works [22, 10, 1]

    RESULTS This section analyzes the speed and accuracy trade-off of the proposed closed-form solution of defocused and spherically- aberrated PSFs. We chose the following methods as base- lines, which have been frequently used in recent computa- tional imaging works [22, 10, 1]. First, we evaluate Hankel transform (Eq. 1) via matrix multiplication and then ...

    work page 2000

  4. [4]

    Ours clearly shows an advantage in computational time, but it can only handle defocus and spherical aberra- tions at the current stage

    DISCUSSION The proposed approximate closed-form PSF evaluation is complementary to the classic Hankel or FFT-based PSF eval- uation. Ours clearly shows an advantage in computational time, but it can only handle defocus and spherical aberra- tions at the current stage. The latter are more suitable for rendering PSFs for custom pupil functions; in constrast...

    work page

  5. [5]

    APPENDIX Full derivation of the closed-form solution.Eq. 4 can be decomposed into the weighted summation of the following six forms of integrals: h1(k)≜ Z a0(k) 0 rΦ(r)dr, h2(k)≜ Z a0(k) 0 r3 Φ(r)dr, h3(k)≜ Z a0(k) 0 r5 Φ(r)dr, h4(k)≜ Z R a0(k) C(r, k) Φ(r;k)dr, h5(k)≜ Z R a0(k) r C(r, k) Φ(r;k)dr, h6(k)≜ Z R a0(k) r2 C(r, k) Φ(r;k)dr. (12) Ours (57.21 s)...

    work page

  6. [6]

    Depth from coupled optical differentiation: J. luo et al.,

    Junjie Luo, Yuxuan Liu, Emma Alexander, and Qi Guo, “Depth from coupled optical differentiation: J. luo et al.,”In- ternational Journal of Computer Vision, vol. 133, no. 11, pp. 8109–8126, 2025

    work page 2025

  7. [7]

    Blurry-edges: Photon-limited depth estimation from defo- cused boundaries,

    Wei Xu, Charles James Wagner, Junjie Luo, and Qi Guo, “Blurry-edges: Photon-limited depth estimation from defo- cused boundaries,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 432–441

    work page 2025

  8. [8]

    Phase- cam3d—learning phase masks for passive single view depth estimation,

    Yicheng Wu, Vivek Boominathan, Huaijin Chen, Aswin Sankaranarayanan, and Ashok Veeraraghavan, “Phase- cam3d—learning phase masks for passive single view depth estimation,” in2019 IEEE International Conference on Com- putational Photography (ICCP). IEEE, 2019, pp. 1–12

    work page 2019

  9. [9]

    Stripformer: Strip transformer for fast im- age deblurring,

    Fu-Jen Tsai, Yan-Tsung Peng, Yen-Yu Lin, Chung-Chi Tsai, and Chia-Wen Lin, “Stripformer: Strip transformer for fast im- age deblurring,” inEuropean conference on computer vision. Springer, 2022, pp. 146–162

    work page 2022

  10. [10]

    Efficient frequency domain-based transform- ers for high-quality image deblurring,

    Lingshun Kong, Jiangxin Dong, Jianjun Ge, Mingqiang Li, and Jinshan Pan, “Efficient frequency domain-based transform- ers for high-quality image deblurring,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2023, pp. 5886–5895

    work page 2023

  11. [11]

    Extended depth of focus imaging using optics and image pro- cessing,

    Chrispin Jiji, VL Nandhini, VL Malini, and R Ciro Rodriguez, “Extended depth of focus imaging using optics and image pro- cessing,”International Journal of Information Technology, vol. 16, no. 2, pp. 1137–1143, 2024

    work page 2024

  12. [12]

    Algorithms for render- ing depth of field effects in computer graphics,

    Brian A Barsky and Todd J Kosloff, “Algorithms for render- ing depth of field effects in computer graphics,” inProceed- ings of the 12th WSEAS international conference on Comput- ers. World Scientific and Engineering Academy and Society (WSEAS), 2008, vol. 2008

    work page 2008

  13. [13]

    Fast depth of field rendering with surface splatting,

    Jaroslav Kriv ´anek, Jiri Zara, and Kadi Bouatouch, “Fast depth of field rendering with surface splatting,” inProceedings Com- puter Graphics International 2003. IEEE, 2003, pp. 196–201

    work page 2003

  14. [14]

    End-to-end optimization of optics and image processing for achromatic extended depth of field and super- resolution imaging,

    Vincent Sitzmann, Steven Diamond, Yifan Peng, Xiong Dun, Stephen Boyd, Wolfgang Heidrich, Felix Heide, and Gordon Wetzstein, “End-to-end optimization of optics and image processing for achromatic extended depth of field and super- resolution imaging,”ACM Transactions on Graphics (TOG), vol. 37, no. 4, pp. 1–13, 2018

    work page 2018

  15. [15]

    Beating spectral bandwidth limits for large aperture broadband nano- optics,

    Johannes E Fr ¨och, Praneeth Chakravarthula, Jipeng Sun, Ethan Tseng, Shane Colburn, Alan Zhan, Forrest Miller, Anna Wirth- Singh, Quentin AA Tanguy, Zheyi Han, et al., “Beating spectral bandwidth limits for large aperture broadband nano- optics,”Nature communications, vol. 16, no. 1, pp. 3025, 2025

    work page 2025

  16. [16]

    Neural nano-optics for high-quality thin lens imaging,

    Ethan Tseng, Shane Colburn, James Whitehead, Luocheng Huang, Seung-Hwan Baek, Arka Majumdar, and Felix Heide, “Neural nano-optics for high-quality thin lens imaging,”Na- ture communications, vol. 12, no. 1, pp. 6493, 2021

    work page 2021

  17. [17]

    Computational imaging through atmospheric turbulence,

    Stanley H Chan, Nicholas Chimitt, et al., “Computational imaging through atmospheric turbulence,”F oundations and Trends® in Computer Graphics and Vision, vol. 15, no. 4, pp. 253–508, 2023

    work page 2023

  18. [18]

    Goodman,F ourier Optics, McGraw-Hill, New York, 2 edi- tion, 1996

    J. Goodman,F ourier Optics, McGraw-Hill, New York, 2 edi- tion, 1996

    work page 1996

  19. [19]

    A differentiable wave op- tics model for end-to-end computational imaging system opti- mization,

    Chi-Jui Ho, Yash Belhe, Steve Rotenberg, Ravi Ramamoorthi, Tzu-Mao Li, and Nicholas Antipa, “A differentiable wave op- tics model for end-to-end computational imaging system opti- mization,” inProceedings of the IEEE/CVF International Con- ference on Computer Vision, 2025, pp. 28042–28051

    work page 2025

  20. [20]

    Successive optimization of optics and post-processing with differentiable coherent psf operator and field information,

    Zheng Ren, Jingwen Zhou, Wenguan Zhang, Jiapu Yan, Bingkun Chen, Huajun Feng, and Shiqi Chen, “Successive optimization of optics and post-processing with differentiable coherent psf operator and field information,”IEEE Transac- tions on Computational Imaging, 2025

    work page 2025

  21. [21]

    Differentiable compound optics and processing pipeline op- timization for end-to-end camera design,

    Ethan Tseng, Ali Mosleh, Fahim Mannan, Karl St-Arnaud, Avinash Sharma, Yifan Peng, Alexander Braun, Derek Nowrouzezahrai, Jean-Francois Lalonde, and Felix Heide, “Differentiable compound optics and processing pipeline op- timization for end-to-end camera design,”ACM Transactions on Graphics (TOG), vol. 40, no. 2, pp. 1–19, 2021

    work page 2021

  22. [22]

    Learning lens blur fields,

    Esther YH Lin, Zhecheng Wang, Rebecca Lin, Daniel Miau, Florian Kainz, Jiawen Chen, Xuaner Zhang, David B Lindell, and Kiriakos N Kutulakos, “Learning lens blur fields,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    work page 2025

  23. [23]

    Accel- erating atmospheric turbulence simulation via learned phase- to-space transform,

    Zhiyuan Mao, Nicholas Chimitt, and Stanley H Chan, “Accel- erating atmospheric turbulence simulation via learned phase- to-space transform,” inProceedings of the IEEE/CVF interna- tional conference on computer vision, 2021, pp. 14759–14768

    work page 2021

  24. [24]

    Basic wavefront aber- ration theory for optical metrology,

    James C Wyant and Katherine Creath, “Basic wavefront aber- ration theory for optical metrology,”Applied optics and optical engineering, vol. 11, no. part 2, pp. 28–39, 1992

    work page 1992

  25. [25]

    Uncorrelated modes of the non-linear power spectrum,

    AJS Hamilton, “Uncorrelated modes of the non-linear power spectrum,”Monthly Notices of the Royal Astronomical Society, vol. 312, no. 2, pp. 257–284, 2000

    work page 2000

  26. [26]

    Abramowitz and I

    M. Abramowitz and I. A. Stegun,Handbook of Mathematical Functions with F ormulas, Graphs, and Mathematical Tables, U.S. Dept. of Commerce, 9 edition, 1964

    work page 1964

  27. [27]

    D-flat: A differentiable flat-optics framework for end-to-end metasurface visual sensor design,

    Dean S Hazineh, Soon Wei Daniel Lim, Zhujun Shi, Federico Capasso, Todd Zickler, and Qi Guo, “D-flat: A differentiable flat-optics framework for end-to-end metasurface visual sensor design,”arXiv preprint arXiv:2207.14780, 2022

    work page arXiv 2022