pith. sign in

arxiv: 2606.28136 · v1 · pith:2AGBZP2Knew · submitted 2026-06-26 · 🌌 astro-ph.IM · cs.CV· physics.optics

Differentiable design of the PIAA-ZWFS: a flexible wavefront sensor that approaches the fundamental limit

A. K. Taras , S. Y. Haffert , L. Desdoigts This is my paper

Pith reviewed 2026-06-29 02:14 UTC · model grok-4.3

classification 🌌 astro-ph.IM cs.CVphysics.optics
keywords wavefront sensoradaptive opticsZernikeapodisationPIAAextreme AOdifferentiable designfundamental limit
0
0 comments X

The pith

The PIAA-ZWFS approaches the fundamental limit of wavefront sensing by a factor of ten over conventional designs.

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

The paper develops the phase-induced amplitude apodisation Zernike wavefront sensor to improve extreme adaptive optics for high contrast astronomy. It applies lossless apodisation to concentrate starlight and uses a differentiable model to optimize the sensor for lower variance in wavefront estimation. The result is better performance across different conditions, including extended sources, without losing dynamic range. A proof shows that all wavefront sensors face a trade-off between amplitude and phase information.

Core claim

The PIAA-ZWFS is an adaptation of the Zernike wavefront sensor that uses phase-induced amplitude apodisation of the pupil to concentrate starlight. Optimized with a differentiable framework to minimize the variance of a maximum likelihood estimator in the high Strehl regime, it closes the gap to the fundamental limit by a factor of 10 compared to the conventional ZWFS in photon-limited cases and by 2.5 compared to an optimized ZWFS. It also outperforms point source sensors for stellar diameters larger than 0.8 λ/D and maintains dynamic range with linear and non-linear reconstructors.

What carries the argument

The PIAA-ZWFS architecture, which combines phase-induced amplitude apodisation with the Zernike wavefront sensor and is optimized through differentiable modeling to minimize estimator variance.

If this is right

  • Closes the gap to the fundamental limit by a factor of 10 compared to conventional ZWFS in typical photon-limited cases.
  • Outperforms the optimised ZWFS by a factor of 2.5 in the same scenario.
  • Outperforms ideal point source sensors for source sizes larger than 0.8 λ/D.
  • Maintains performance with linear or non-linear reconstructors without loss of dynamic range.
  • There is a necessary trade-off between amplitude and phase error information for any wavefront sensor.

Where Pith is reading between the lines

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

  • The sensor could allow adaptive optics systems to run on fainter targets or at higher frame rates for exoplanet detection.
  • Optimization methods using differentiable models may apply to other types of wavefront sensors.
  • Implementation on real telescopes would test if the simulated gains hold under actual atmospheric and instrumental conditions.

Load-bearing premise

The maximum likelihood estimator operates in the high Strehl regime and the differentiable model accurately captures the physical behavior of the PIAA-ZWFS.

What would settle it

An experiment measuring the estimation variance for the PIAA-ZWFS in a photon-limited, high Strehl setup and checking if it matches the predicted improvement over standard ZWFS.

Figures

Figures reproduced from arXiv: 2606.28136 by A. K. Taras, L. Desdoigts, S. Y. Haffert.

Figure 1
Figure 1. Figure 1: Architecture of the PIAA-ZWFS and our optimisation framework. We simulate starlight from a point or resolved source [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Verification that PIAA-ZWFS can reach the sensitivity [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: PIAA ZWFS designs and behaviour for different aper￾tures. The apodization profile after the first lens pair (1st row, showing |E|) concentrates the light in the centre of the pupil. The reference intensity (2nd row, now at the sampling of the detector) scatters some light outside the pupil. The intensity change for an applied Fourier mode (3rd and 4th rows) shows a strong response, however if the inverse P… view at source ↗
Figure 4
Figure 4. Figure 4: PIAA-ZWFS performance on the ELT (240 pixels across [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Resolved objects on the ELT, Z band. Top: WFS perfor￾mance as a function of stellar size. Performance declines with coherence, notably for the ideal WFS (black, dashed) which meets the quantum information (QI) bound (black dot) for point sources. Our optimised designs for resolved objects (orange and red) outperform the “ideal” system for D⋆ = 0.8λ/D. Middle: Information (relative to the point source limit… view at source ↗
Figure 7
Figure 7. Figure 7: Wavefront estimate errors under varying input aberra [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
read the original abstract

Extreme adaptive optics (AO) is necessary for high contrast astronomy at scales of the habitable zone of nearby systems. We seek to evaluate wavefront sensors that approach fundamental limits of wavefront sensing, enabling adaptive optics systems to run faster or on fainter targets. We present the phase-induced amplitude apodisation Zernike wavefront sensor (PIAA-ZWFS): an adaptation of the conventional Zernike wavefront sensor (ZWFS) that leverages lossless apodisation of the pupil to concentrate the starlight in the focal plane. We optimise and evaluate the sensor with a differentiable modelling framework, drawing on concepts from Bayesian experimental design to minimise the variance of a maximum likelihood estimator that uses the system in the high Strehl regime. Our architecture shows state-of-the-art performance in simulation for different apertures, bandwidths, photon fluxes and source sizes, closing the gap to the fundamental limit by a factor 10 (2.5) compared to the conventional ZWFS (optimised ZWFS) in a typical photon-limited case. For extended sources, we show that even an ideal point source sensor rapidly becomes sub-optimal, and our system outperforms it for stellar diameters larger than 0.8{\lambda}/D. We verify that these gains do not come at the cost of dynamic range with either linear or non-linear reconstructors. Finally, we present a proof that there must be a trade-off between the information gained about amplitude and phase errors for any wavefront sensor. The PIAA-ZWFS is a viable wavefront sensor operating near the fundamental sensitivity limits.

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 / 3 minor

Summary. The manuscript introduces the phase-induced amplitude apodisation Zernike wavefront sensor (PIAA-ZWFS), an adaptation of the conventional Zernike wavefront sensor (ZWFS) that uses lossless pupil apodisation to concentrate starlight. A differentiable physical model is employed to optimise the design by minimising the variance of a maximum likelihood estimator (MLE) in the high Strehl regime, drawing on Bayesian experimental design concepts. Simulations across apertures, bandwidths, photon fluxes and source sizes demonstrate state-of-the-art performance, closing the gap to the fundamental limit by a factor of 10 (2.5) relative to the conventional ZWFS (optimised ZWFS) in a typical photon-limited case. Additional results address extended sources (outperforming point-source sensors for diameters >0.8 λ/D), dynamic range with linear/non-linear reconstructors, and a proof of the necessary amplitude-phase information trade-off for any wavefront sensor.

Significance. If the simulation results hold under the stated modeling assumptions, the PIAA-ZWFS represents a meaningful advance for extreme adaptive optics in high-contrast astronomy, potentially enabling faster correction or fainter targets. The differentiable optimisation framework is a methodological strength that could generalise to other sensor designs, and the theoretical proof on information trade-offs adds lasting value. The work is entirely simulation-based with no experimental validation or independent model benchmarking reported.

major comments (2)
  1. [Results / optimisation section] The central performance claims (factor-of-10 and factor-of-2.5 gap closures) rest on MLE variance comparisons to a fundamental limit; the manuscript must provide explicit equations or tables (e.g., in the results section) showing how the limit is computed, how the MLE variance is evaluated, and the precise Strehl ratios at which the high-Strehl assumption is applied, as these quantities are load-bearing for the quantitative claims.
  2. [Methods / simulation framework] The differentiable model is used both to optimise the PIAA-ZWFS and to evaluate its performance; an independent cross-check (e.g., against a non-differentiable physical optics code or analytic ZWFS expressions) is required to confirm that optimisation gains are not artefacts of the shared model, particularly for the extended-source and dynamic-range cases.
minor comments (3)
  1. Clarify the exact definition and derivation of the 'fundamental limit' used for benchmarking (abstract and § on information limits).
  2. The dynamic-range verification with linear and non-linear reconstructors is mentioned but lacks specifics on the reconstructors or the quantitative metric; add a short table or figure caption with these details.
  3. Notation for the PIAA apodisation parameters and the MLE covariance should be defined consistently in the first methods subsection where they appear.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their positive assessment and constructive comments, which will strengthen the manuscript. We address each major comment below and will revise accordingly.

read point-by-point responses

  1. Referee: [Results / optimisation section] The central performance claims (factor-of-10 and factor-of-2.5 gap closures) rest on MLE variance comparisons to a fundamental limit; the manuscript must provide explicit equations or tables (e.g., in the results section) showing how the limit is computed, how the MLE variance is evaluated, and the precise Strehl ratios at which the high-Strehl assumption is applied, as these quantities are load-bearing for the quantitative claims.

    Authors: We agree that explicit documentation of these quantities is necessary to support the quantitative claims. In the revised manuscript we will add, in the results section, the explicit equations for the fundamental limit (Cramér-Rao bound under the high-Strehl approximation), the procedure used to evaluate MLE variance (via the inverse Fisher information matrix), and a table listing the precise Strehl ratios (all simulations use Strehl > 0.9) at which the high-Strehl assumption is invoked. revision: yes

  2. Referee: [Methods / simulation framework] The differentiable model is used both to optimise the PIAA-ZWFS and to evaluate its performance; an independent cross-check (e.g., against a non-differentiable physical optics code or analytic ZWFS expressions) is required to confirm that optimisation gains are not artefacts of the shared model, particularly for the extended-source and dynamic-range cases.

    Authors: We acknowledge the value of independent validation. In revision we will add, in the methods section, direct comparisons of the differentiable model against analytic ZWFS expressions for the conventional (non-PIAA) case across the reported photon fluxes and bandwidths. For the PIAA-ZWFS itself, where no closed-form solution exists, we will include additional notes on the physical-optics derivation of the model and a limited cross-check against a non-differentiable Fourier-propagation implementation for the extended-source and dynamic-range results. Full benchmarking of every optimized configuration against an independent code base is not feasible within the current scope but the requested analytic checks will be provided. revision: partial

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper optimizes the PIAA-ZWFS via a differentiable physical model to minimize MLE variance (high-Strehl regime) using Bayesian experimental design concepts, then evaluates simulated performance against an independent fundamental limit (Cramér-Rao type) across apertures, bandwidths, fluxes and source sizes. A separate proof addresses amplitude-phase information trade-off. No quoted equations or steps reduce by construction to fitted inputs, self-citations, or renamed known results; the modeling accuracy is an explicit assumption rather than a derived claim. Central results remain externally falsifiable via simulation benchmarks and the stated limit, qualifying as self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claims rest on the domain assumption of high Strehl regime operation for the MLE and the fidelity of the differentiable simulation model; no explicit free parameters are identified.

axioms (1)
  • domain assumption The maximum likelihood estimator uses the system in the high Strehl regime
    Stated explicitly in the abstract as the operating context for the estimator.
invented entities (1)
  • PIAA-ZWFS no independent evidence
    purpose: Wavefront sensor that approaches the fundamental limit via lossless apodisation
    New design introduced as an adaptation of the conventional ZWFS.

pith-pipeline@v0.9.1-grok · 5834 in / 1364 out tokens · 47787 ms · 2026-06-29T02:14:11.759756+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

33 extracted references · 5 canonical work pages

  1. [1]

    & Kahn, P

    Bracewell, R. & Kahn, P. B. 1966, American Journal of Physics, 34, 712

    1966

  2. [2]

    2022, in Adaptive Optics Systems VIII, V ol

    Chambouleyron, V ., Fauvarque, O., Plantet, C., et al. 2022, in Adaptive Optics Systems VIII, V ol. 12185, SPIE, 932–944

    2022

  3. [3]

    2023, Astronomy & Astro- physics, 670, A153

    Chambouleyron, V ., Fauvarque, O., Plantet, C., et al. 2023, Astronomy & Astro- physics, 670, A153

    2023

  4. [4]

    K., Jensen-Clem, R., & Macintosh, B

    Chambouleyron, V ., Wallace, J. K., Jensen-Clem, R., & Macintosh, B. 2024, Optics Express, 32, 47706

    2024

  5. [5]

    2018, Annual Review of Materials Research, 48, 279

    Chang, S., Guo, X., & Ni, X. 2018, Annual Review of Materials Research, 48, 279

    2018

  6. [6]

    2025, Astronomy & Astro- physics, 701, A157

    Darcis, M., Haffert, S., Chambouleyron, V ., et al. 2025, Astronomy & Astro- physics, 701, A157

    2025

  7. [7]

    S., Fagginger Auer, F., Escuti, M

    Doelman, D. S., Fagginger Auer, F., Escuti, M. J., & Snik, F. 2018, Optics letters, 44, 17

    2018

  8. [8]

    S., Snik, F., Warriner, N

    Doelman, D. S., Snik, F., Warriner, N. Z., & Escuti, M. J. 2017, in Techniques and Instrumentation for Detection of Exoplanets VIII, V ol. 10400, SPIE, 224–235

    2017

  9. [9]

    2019, Journal of the Optical Society of America A, 36, 1241

    Fauvarque, O., Janin-Potiron, P., Correia, C., et al. 2019, Journal of the Optical Society of America A, 36, 1241

    2019

  10. [10]

    2013, Practical methods of optimization (John Wiley & Sons)

    Fletcher, R. 2013, Practical methods of optimization (John Wiley & Sons)

    2013

  11. [11]

    Forbes, G. W. 2007, Optics express, 15, 5218

    2007

  12. [12]

    2019, Advanced Optical Technologies, 8, 163

    Grushina, A. 2019, Advanced Optical Technologies, 8, 163

    2019

  13. [13]

    2003, Astronomy & Astrophysics, 404, 379

    Guyon, O. 2003, Astronomy & Astrophysics, 404, 379

    2003

  14. [14]

    2005, The Astrophysical Journal, 629, 592

    Guyon, O. 2005, The Astrophysical Journal, 629, 592

    2005

  15. [15]

    2018, Annual Review of Astronomy and Astrophysics, 56, 315

    Guyon, O. 2018, Annual Review of Astronomy and Astrophysics, 56, 315

    2018

  16. [16]

    Y ., Males, J

    Haffert, S. Y ., Males, J. R., & Guyon, O. 2023, arXiv preprint arXiv:2310.10889

    work page arXiv 2023

  17. [17]

    Hoffnagle, J. A. & Jefferson, C. M. 2000, Applied optics, 39, 5488

    2000

  18. [18]

    C., Pathak, P., et al

    Kasper, M., Urra, N. C., Pathak, P., et al. 2021, arXiv preprint arXiv:2103.11196

    work page arXiv 2021

  19. [19]

    Y ., Long, J., et al

    Landman, R., Haffert, S. Y ., Long, J., et al. 2025, Astronomy & Astrophysics, 696, L1

    2025

  20. [20]

    H., et al

    Landman, R., Keller, C., Por, E. H., et al. 2022, in Adaptive Optics Systems VIII, V ol. 12185, SPIE, 2582–2593

    2022

  21. [21]

    R., Close, L

    Males, J. R., Close, L. M., Haffert, S. Y ., et al. 2024, in Ground-based and Air- borne Instrumentation for Astronomy X, ed. J. R. Vernet, J. J. Bryant, & K. Motohara (Yokohama, Japan: SPIE), 34

    2024

  22. [22]

    R., Fitzgerald, M

    Males, J. R., Fitzgerald, M. P., Belikov, R., & Guyon, O. 2021, Publications of the Astronomical Society of the Pacific, 133, 104504

    2021

  23. [23]

    A., Bailey, J., Bartlett, J., et al

    Mazin, B. A., Bailey, J., Bartlett, J., et al. 2019, arXiv preprint arXiv:1908.02775 N’Diaye, M., Dohlen, K., Fusco, T., & Paul, B. 2013, Astronomy & Astro- physics, 555, A94

    work page arXiv 2019

  24. [24]

    2008, 139, 012021

    Paterson, C. 2008, 139, 012021

    2008

  25. [25]

    Rader, T

    Rader, J., Lyons, T., & Kidger, P. 2024, arXiv preprint arXiv:2402.09983

    work page arXiv 2024

  26. [26]

    & Farinato, J

    Ragazzoni, R. & Farinato, J. 1999, Astronomy and Astrophysics, v. 350, p. L23- L26 (1999), 350, L23

    1999

  27. [27]

    1965, Journal of the Optical Society of America, 55, 1110

    Slepian, D. 1965, Journal of the Optical Society of America, 55, 1110

    1965

  28. [28]

    Soummer, R., Pueyo, L., Sivaramakrishnan, A., & Vanderbei, R. J. 2007, Optics Express, 15, 15935

    2007

  29. [29]

    K., Kettenbeil, C., & Jewell, J

    Steeves, J., Wallace, J. K., Kettenbeil, C., & Jewell, J. 2020, Optica, 7, 1267

    2020

  30. [30]

    & Ashok, A

    Trzaska, J. & Ashok, A. 2025, arXiv preprint arXiv:2511.05707

    work page arXiv 2025

  31. [31]

    & Wallace, J

    Wenger, T. & Wallace, J. K. 2025, Optics Letters, 50, 726

    2025

  32. [32]

    1979, Optics Communications, 28, 21

    Wyne, C. 1979, Optics Communications, 28, 21

    1979

  33. [33]

    |C {WkEi}|2 dudv − ⟨C {iWkEi} ,C {Ei}⟩ ⟨C {Ei} ,C {iWkEi}⟩ .(A.35) Similarly for the amplitude-amplitude term FIQ bkbk =4ℜ

    Zernike, F. 1934, Monthly Notices of the Royal Astronomical Society, V ol. 94, p. 377-384, 94, 377 Article number, page 10 A. K. Taras et al.: Differentiable design of the PIAA-ZWFS: a flexible wavefront sensor that approaches the fundamental limit Appendix A: Derivation of Fisher information bounds A general result for the derivative of the intensity wit...

    1934