Pith. sign in

REVIEW 4 major objections 5 minor 28 references

Enhanced interferometric resolution via N-fold intensity-product measurements without sacrificing phase sensitivity

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that multiplying the intensities from N equal divisions of a Mach-Zehnder output narrows the central interference fringe, with full width at half maximum shrinking as 1/√N, while the phase sensitivity stays identical to…

desk verdict The fringe narrowing is a mathematical identity, the phase-sensitivity claim is built on a wrong Fisher information formula, but the experimental demonstration is honest and the paper is worth a critical referee. read the letter →

arxiv 2506.06061 v1 pith:C7HLMIWC submitted 2025-06-06 physics.optics

S. Kim , J. Stohr , B. S. Ham This is my paper
classification physics.optics
keywords intensity-productmeasurementMach-ZehnderinterferometerfringeresolutionphasesensitivityFisherinformationshot-noiselimitPoissonstatisticssuperresolution
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper reports a classical, readily implementable way to make the interference fringes of a Mach-Zehnder interferometer sharper: split one output into N equal subfields, measure the intensity in each arm, and multiply the N readings together. The paper's organizing distinction is that resolution (fringe width) and phase sensitivity (minimum detectable phase step) are separate quantities, so a sharper central fringe need not cost sensitivity. Near a phase of zero the resulting Nth-order intensity product, $G^{(N)}(\phi)=I_0^N(1+\cos\phi)^N$, has a central peak whose full width at half maximum narrows as $1/\sqrt{N}$, a shot-noise-limit-like scaling usually associated with quantum resources. The paper argues through a Fisher-information calculation that this narrowing costs nothing in phase sensitivity, leaving the minimum detectable phase step the same as that of the standard single-output interferogram. The authors demonstrate the effect experimentally for $N=2,3,4$ in both single-photon coincidence counting and continuous-wave detection, and they are explicit that the central-fringe gain is local, with a compensating broadening near $\phi=\pm\pi$ and no net resolution gain over a full fringe period.

What carries the argument

The central object is the $N$th-order intensity-correlation function $G^{(N)}(\phi)=\prod_{k=1}^N I_k(\phi)=I_0^N(1+\cos\phi)^N$, built by splitting one Mach-Zehnder output into N arms and multiplying the detected intensities. Its role is to reshape the interferogram: near $\phi=0$ the factor $(1+\cos\phi)^N$ behaves like $2^N e^{-N\phi^2/4}$, which is why the central FWHM contracts as $1/\sqrt{N}$. The accompanying analysis uses the Fisher-information expression $F(\phi)=(\partial_\phi\langle X\rangle)^2/\mathrm{Var}(X)$ for the product observable $X=\prod_k I_k$, with the N subfields assumed to be independent, identically distributed Poisson variables; this is the step that produces the $N$-independent phase sensitivity $\Delta\phi=1/\sqrt{2I_0}$ quoted for the standard MZI output.

What would settle it

Simulate N independent Poisson detectors with mean counts $I_0(1+\cos\phi)/N$, form their product, and compute the true minimum phase uncertainty from the Poisson likelihood on a fine grid of $\phi$; if that uncertainty depends on N, the paper's claim of preserved, N-independent phase sensitivity fails, and if it is N-independent, the claim stands.

Watch

Extended reading notes

Core claim

The central discovery is that the $N$-fold intensity product of equally divided Mach-Zehnder output subfields, $G^{(N)}(\phi)=I_0^N(1+\cos\phi)^N$, sharpens the constructive-interference fringe while leaving the phase sensitivity at the shot-noise limit. With the N subfields treated as independent and identically distributed Poisson observables, the Fisher information of the product is argued to be $F^{(N)}(\phi)=(\partial_\phi\langle G\rangle)^2/\mathrm{Var}(G)$, and the calculation yields a minimum phase uncertainty $\Delta\phi=1/\sqrt{2I_0}$ that is independent of N and identical to the standard MZI output. The paper confirms the predicted $\pi/\sqrt{N}$ scaling of the central FWHM experimentally for $N=2,3,4$, using either single-photon counters with coincidence post-selection or ordinary fast photodetectors on a continuous-wave beam, and shows that the same product approach applied to the two output ports doubles the number of fringes and adds an extra $\sqrt{2}$ in the continuous-wave regime. The paper also states the trade-off explicitly: the enhanced resolution near $\phi=0$ is offset by a resolution loss near $\phi=\pm\pi$, so the global resolution over the full phase period is not improved.

Load-bearing premise

The claim that phase sensitivity is preserved assumes that the Fisher information of the product of N Poisson-distributed intensity readings is exactly the slope squared divided by the variance; that identity holds strictly only for Gaussian observables, so the N-independence of the phase uncertainty is not guaranteed by the calculation as written.

Editorial extensions

If this is right

  • Any conventional Mach-Zehnder-based sensor, such as a fiber-optic gyroscope or wavelength meter, can adopt the technique by inserting a beam-splitter tree and multiplying detector outputs, with no change to the light source or its power.
  • The $\sqrt{N}$ central-fringe narrowing is available in both the single-photon coincidence regime and the continuous-wave regime, so it does not require quantum light or coincidence detectors in the high-power setting.
  • Using the two MZI output ports together in a correlated product doubles the fringe count and adds an extra $\sqrt{2}$ to the resolution gain in the continuous-wave regime.
  • The phase sensitivity of the product signal remains at the shot-noise limit set by the input intensity, so the resolution gain does not come at the cost of sensitivity.
  • Because the enhancement is local, the practical operating point is near the sharpened central fringe; near $\phi=\pm\pi$ the fringes broaden, so the method is not a uniform improvement over the full period.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Computing the exact Fisher information of the product observable from the Poisson likelihood of the N subfield counts would settle whether the quoted $N$-independent $\Delta\phi$ is exact or only a bound, since the slope-squared-over-variance formula used in the paper is exact only for Gaussian observables.
  • The same splitting-and-multiplying recipe should transfer to any interferometric sensor with a sinusoidal response, such as a ring cavity, Fabry-Pérot cavity, or grating spectrometer, provided its output field can be evenly divided before detection.
  • The local nature of the gain suggests an engineering use: operate near the sharpened central fringe and use the steep side of the product fringe as a high-slope discriminant for wavelength or displacement tracking, accepting that the response must be re-referenced across fringe periods.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The manuscript proposes an N-fold intensity-product readout of a Mach-Zehnder interferometer: one output port is split into N subfields and the product G^(N)(φ) = I0^N (1+cosφ)^N (Eq. 4) is used as the signal. The authors show, numerically and experimentally, that the central-fringe FWHM of this product signal scales as 1/√N, and they interpret this as a shot-noise-limit-like resolution enhancement. From a Fisher-information calculation in the Discussion (Eqs. 5–6), they claim that the phase sensitivity of the N-fold product is N-independent and equal to that of the standard MZI output; a cross-product of the two MZI output ports is analyzed similarly (Eqs. 7–8). The conclusion is that intensity-product measurements enhance interferometric resolution without sacrificing phase sensitivity.

Significance. If the Fisher-information claim were correct, this would be a simple, fully classical method for improving interferometric fringe resolution at fixed input power, with potential applications in fiber-optic gyroscopes and wavelength meters. The experimental data do confirm that the FWHM of the averaged product signal follows the predicted 1/√N behavior, and the paper usefully distinguishes fringe resolution (FWHM) from phase sensitivity (slope-to-noise ratio). However, the central sensitivity claim rests on a nonstandard and, for this problem, incorrect identification of Fisher information with (∂⟨X⟩/∂φ)²/Var(X). The FWHM narrowing is a mathematical property of taking powers of a sinusoidal fringe, and at the very phase point where the narrowing is largest the slope—and, under the authors' own formula, the Fisher information—vanishes. As it stands, the paper does not establish any gain in estimation precision, and the title claim is not supported.

major comments (4)
  1. [Discussion: Fisher information, Eq. (5)] The formula F(X) = (∂⟨X⟩/∂φ)² / Var(X) is not the definition of Fisher information for an arbitrary observable; it is the Fisher information for a Gaussian model or, more generally, for an exponential family in which X is the canonical sufficient statistic. For G = ∏ X_j with X_j independent Poisson variables, the true Fisher information is E[(∂ log p_G(G;φ)/∂φ)²], not the delta-method expression in Eq. (5). The quantity in Eq. (5) is a lower bound on the true Fisher information of G, not the Fisher information itself. Therefore Eq. (6), Δφ_N = 1/√[I0(1−cosφ)], is not the Cramér-Rao lower bound for estimating φ from the product measurement, and the claim that the product preserves the standard MZI phase sensitivity is unsupported.
  2. [Discussion, Eqs. (5)–(6); Fig. 2] The claimed equality FI_G = I0(1−cosφ) contradicts the data-processing inequality. The joint count vector (X_1,...,X_N) from the N subfields has Fisher information Σ_j (∂μ_j/∂φ)²/μ_j = I0(1−cosφ) at fixed total power, and G is a deterministic function of this vector. Hence FI_G ≤ I0(1−cosφ), with strict inequality for N > 1 because the product is not the sufficient statistic (the sum Σ X_j is). Thus the product observable generically has strictly lower Fisher information than the standard MZI count measurement, not identical to it. The statement in the Discussion that the phase sensitivity of the N-fold intensity product remains identical to that of the standard MZI output is therefore not only unproven but in conflict with a basic information-theoretic inequality.
  3. [Results, Eq. (4) and Fig. 2] The FWHM narrowing near φ=0 is an algebraic property of (1+cosφ)^N and does not by itself constitute an enhancement of estimation capability. At φ=0 the derivative of G^(N)(φ) is zero for N > 1, and the authors' own Eq. (5) gives F^(N)(0)=0, so the phase uncertainty Δφ diverges exactly where the FWHM is minimized. The claim that this behavior is 'not anticipated by classical Fisher information' is misleading: classical Fisher information predicts that the central fringe, however narrow in the mean, carries no phase information at its peak. The resolution-versus-sensitivity discussion in the paper does not resolve this issue because the proposed sensing application relies on locating or estimating a phase from the fringe shape.
  4. [Discussion, Eq. (5), variance evaluation; Methods] The variance evaluation in Eq. (5) uses the large-μ approximation (1+μ)^N − μ^N ≈ N μ^{N−1}, which is only valid for μ ≫ 1. The exact variance of the product of N independent Poisson variables is (μ+μ²)^N − μ^{2N}, which contains additional terms at finite μ. The claimed N-independence of Δφ is therefore not established in the single-photon regime of Fig. 3(a), where the mean photon number per subfield is not large; the same defect affects the cross-product analysis in Eqs. (7)–(8).
minor comments (5)
  1. [Fig. 2] The right panel does not define the plotted red curves or the normalization used for the FWHM values; please provide analytic expressions for the FWHM of (1+cosφ)^N near φ=0 and near φ=±π.
  2. [Equations (1)–(8)] The displayed equations contain numerous typesetting errors and missing symbols, particularly in Eqs. (1)–(8); the manuscript needs careful proofreading before any resubmission.
  3. [References] Reference [14] is a duplicate of reference [3]; the reference list should be consolidated and checked for completeness.
  4. [Methods] The Methods state that the mean photon count per data point is of order 10^?; the exponent is illegible in the submitted text and should be stated explicitly.
  5. [Introduction] The statement that the order N is limited by the photon number of the input light is not meaningful for a CW source with ~10^13 photons/s; in practice N is limited by the number of detectors and by the detection electronics, not by the photon flux.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the resolution scaling is a direct mathematical consequence of the product observable, the phase-sensitivity result follows algebraically from the stated model, and the self-citations are background rather than load-bearing.

full rationale

The paper's derivation chain is self-contained. Equation (4), G^(N)(φ)=I0^N(1+cosφ)^N, is not fitted or inferred from the later claims; it is the definition of the intensity-product observable built from N equally divided MZI subfields under the stated Poisson/i.i.d. assumption. The FWHM scaling of (1+cosφ)^N as 1/√N is a mathematical property of that definition, and Fig. 2 is a numerical evaluation of the same formula. The subsequent experiment in Fig. 3 is an independent test of whether the physical apparatus realizes that functional form. The Fisher-information calculation in Eq. (5) also follows algebraically from the same model (mean and variance of the product statistic in the large-μ approximation), and the N-independence of Δφ in Eq. (6) is a consequence of that algebra, not of a fitted parameter. The self-citations (refs. 23, 24, 27) are cited as background for the intensity-product method and the cross-intensity generalization; none is invoked as an external uniqueness theorem or as a substitute for the paper's own derivation. There is therefore no circular step in the sense of a definition containing the target result, a fitted input renamed as a prediction, or a load-bearing self-citation chain. (A possible concern is the use of the delta-method expression (∂⟨G⟩/∂φ)^2/Var(G) as the exact Fisher information for a non-Gaussian product observable; that is a correctness issue, not a circularity.)

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No new physical entities are introduced. The central derivation relies on an incorrect Fisher information identity, treated as an assumption, and on the i.i.d. Poisson assumption for the subfields, which holds only in the CW regime. The large-μ approximation is a standard mathematical simplification.

assumptions (3)
  • ad hoc to paper The Fisher information of an observable X is F = (∂⟨X⟩/∂φ)^2 / Var(X).
    Used in the Discussion (Eq. 5) to compute the phase sensitivity of the N-fold product. This identity is only valid for Gaussian observables, not for a product of Poisson variables.
  • domain assumption The N-divided output subfields are independent and identically distributed Poisson random variables.
    Stated in Results and Discussion. True for coherent light split by beam splitters in the CW regime, but questionable for the single-photon post-selected coincidence regime.
  • standard math For large mean μ, (1+1/μ)^N - 1 ≈ N/μ.
    Used to approximate Var(G) in the Fisher information derivation (Discussion).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhanced interferometric resolution via N-fold intensity-product measurements without sacrificing phase sensitivity." pith.science (2026). https://pith.science/paper/C7HLMIWC

@misc{pith2026250606061,
  author       = {Pith},
  title        = {Pith review of: Enhanced interferometric resolution via N-fold intensity-product measurements without sacrificing phase sensitivity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C7HLMIWC}},
  note         = {Machine review of arXiv:2506.06061}
}
read the original abstract

The Fisher information theory sets a fundamental bound on the minimum measurement error achievable from independent and identically distributed (i.i.d.) measurement events. The assumption of identical and independent distribution often implies a Gaussian distribution, as seen in classical scenarios like coin tossing and an optical system exhibiting Poisson statistics. In an interferometric optical sensing platform, this translates to a fundamental limit in phase sensitivity, known as the shot-noise limit (SNL), which cannot be surpassed without employing quantum techniques. Here, we, for the first time to the best of our knowledge, experimentally demonstrate a SNL-like feature on resolution of an unknown signal when intensity-product measurement technique is applied to N-divided MZI output subfields. Given the Poisson-distributed photon statistics, the N-divided subfields ensure the i.i.d. condition required by Fisher information theory. Thus, the N-fold intensity-product technique holds promise for enhancing the precision of conventional optical sensing platforms such as a fiber-optic gyroscope and wavelength meter, while preserving the original phase sensitivity of the output field.

Figures

Figures reproduced from arXiv: 2506.06061 by the authors.

Figure 1
Figure 1. shows the schematic of the intensity product-enhanced sensing metrology in MZI. For this, one MZI output port is evenly divided into four ports for up to fourth-order intensity-product measurements, where measurement scheme of the intensity product between subdivided ports is for either a single-photon or continuous-wave (CW) regime. Importantly, the intensity-product order N is scalable up to the photon number of t… view at source ↗
Figure 2
Figure 2. Numerical calculations of the normalized intensity product. (left panel) For N=1 (dotted), N=2 (blue), N=3 (red), N=4 (green), N=8 (magenta), N=16 (cyan), N=32 (black), N=64 (red-dotted). (right panel) Red (dotted) curve is for 1/√ 2 − 1/√. Circles: FWHM data from the left panel at φ = 0. Stars: FWHM data from the left panel at φ = ±π [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    1 Enhanced interferometric resolution via N-fold intensity-product measurements without sacrificing phase sensitivity S. Kim1, J. Stöhr2, and Byoung S. Ham1,* 1Department of Electrical Engineering and Computer Science, Gwangju Institute of Science and Technology, Gwangju 61005, South Korea 2SLAC National Accelerator Laboratory and Department of Photon Sci...

    work page 2025

  2. [2]

    Ferreira, M. F. S. et al., Roadmap on optical sensors. J. Opt. 19, 083001 (2017)

    work page 2017

  3. [3]

    K., Zhao, G., Wiersig, J

    Chen, W., Özdemir, S. K., Zhao, G., Wiersig, J. and Yang, L. Exceptional points enhance sensing in an optical microcavity. Nature 548, 192-196 (2017)

    work page 2017

  4. [4]

    K., Schmied, R

    Pezzè, L., Smerzi, A., Oberthaler, M. K., Schmied, R. and Treutlein, P. Quantum metrology with nonclassical states of atomic ensembles. Rev. Mod. Phys. 90, 035005 (2018)

    work page 2018

  5. [5]

    McDonagh, C., Burke, C. S. and MacCraith, B. D. Optical chemical sensors. Chem. Rev. 108, 400-422 (2008)

    work page 2008

  6. [6]

    and Diaspro, A

    Vicidomini, G., Bianchini, P. and Diaspro, A. STED super-resolved microscopy. Nature Methods 15, 173-182 (2018)

    work page 2018

  7. [7]

    G., Pitris, C., Boppart, S

    Fujimoto, J. G., Pitris, C., Boppart, S. A. and Brezinski, M. E. Optical coherence tomography: an emerging technology for biomedical imaging and optical biopsy. Neoplasia 2, 9-25 (2000)

    work page 2000

  8. [8]

    P., D’Alessandro, B

    Dhawan, A. P., D’Alessandro, B. and Fu, X. Optical imaging modalities for biomedical applications. IEEE Rev. Bio. Eng. 3, 69-927 (2010)

    work page 2010

Show all 28 references
  1. [9]

    Yao, K. et al. Nanoscale optical imaging of 2D semiconductor stacking orders by exciton-enhanced second harmonic generation. Adv. Optical Mater. 10, 2200085 (2022)

    2022

  2. [10]

    and Lee, M

    Kwon, S., Park, J., Kim, K., Cho, Y. and Lee, M. Microsphere-assisted, nanospot, non-destructive metrology for semiconductor devices. Light: Sci. & Appl. 11, 32 (2022)

    2022

  3. [11]

    K., Pennock, S

    Siddiq, K., Hobden, M. K., Pennock, S. R. and Watson, R. J. Phase noise in FMCW radar systems. IEEE Trans. Aero. Elec. Sys. 55, 70-81 (2019)

    2019

  4. [12]

    and Fascista, A

    Coluccia, A., Parisi, G. and Fascista, A. Detection and classification of multirotor drones in radar sensor networks: a review. Sensors 20, 4172 (2020)

    2020

  5. [13]

    Fundamentals of statistical signal processing: Estimation theory

    Kay, S. Fundamentals of statistical signal processing: Estimation theory. (Prentice Hall PTR, New Jersey 1993)

    1993

  6. [14]

    Tse, M. et al. Quantum-enhanced advanced LIGO detectors in the era of gravitational-wave astronomy. Phys. Rev. Lett. 123, 231107 (2019)

    2019

  7. [15]

    K., Schmied, R

    Pezzè, L., Smerzi, A., Oberthaler, M. K., Schmied, R. and Treutlein, P. Quantum metrology with nonclassical states of atomic ensemble. Rev. Mod. Phys. 90, 035005 (2018)

    2018

  8. [16]

    and Maccone, L

    Giovannetti, V., Lloyd, S. and Maccone, L. Quantum-enhanced measurements: beating the standard quantum limit. Science 306, 1330-1336 (2004)

    2004

  9. [17]

    Large-scale passive laser gyroscope for earth rotation sensing

    Liu, K., et al. Large-scale passive laser gyroscope for earth rotation sensing. Opt. Lett. 44, 2732-2735 (2019)

    2019

  10. [18]

    L., Pedrotti, L

    Pedrotti, F. L., Pedrotti, L. M. and Pedrotti, L. S. Introduction to Optics. 3rd Ed (Pearson Education, Inc. 2007)

    2007

  11. [19]

    Drever, R. W. P., Hall, J. L., Kowalski, F. V., Hough, J., Ford, G. M., Munley, A. J. and Ward, H. Laser phase and frequency stabilization using an optical resonator. Appl. Phys. B. 31, 97-105 (1983)

    1983

  12. [20]

    and Menzel, W

    Saldschmidt ,C., Hasch, J. and Menzel, W. Automative Radar-from first efforts to future systems. IEEE J. Microwaves 1, 135-148 (2021)

    2021

  13. [21]

    Behroozpour, B., Sandborn, P. A. M, Wu, M. and Boser, B. E. Lidar system architectures and circuits. Intg. Circuits. Communi. Mag. 135 (Oct. 2017)

    2017

  14. [22]

    and Kim S

    Sypabekova, M., Hagemann, A., Kleiss, J., Morlan, C. and Kim S. Optimizing an optical cavity-based biosensor for enhanced sensitivity. IEEE Sensors J. 23, 25911 (2023)

    2023

  15. [23]

    GINGERINO: a high sensitivity ring laser gyroscope for fundamental and quantum physics investigation

    Giovinetti, F.; et al. GINGERINO: a high sensitivity ring laser gyroscope for fundamental and quantum physics investigation. Frontiers in Quantum Sci. Tech. 2024, 3,1363409

    2024

  16. [24]

    Overcoming the diffraction limit by multi-photon interference: A tutorial

    Stöhr, J. Overcoming the diffraction limit by multi-photon interference: A tutorial. Adv. Opt. Photonics 11, 215- (2019)

    2019

  17. [25]

    Ham, B. S. Intensity-product-based optical sensing to beat the diffraction limit in an interferometer. Sensors 24, 5041 (2024). 8

    2024

  18. [26]

    Phase shift between the transmitted and the reflected optical fields of a semirefecting lossless mirror is π/2

    Degiorgio, V. Phase shift between the transmitted and the reflected optical fields of a semirefecting lossless mirror is π/2. Am. J. Phys. 48, 81–82 (1980)

    1980

  19. [27]

    Dimitrova, T. L. & Weis, A. The wave-particle duality of light: A demonstration experiment. Am. J. Phys. 76, 137-142 (2008)

    2008

  20. [28]

    Ham, B. S. A superresolution-enhanced spectrometer beyond the Cramer-Rao bound in phase sensitivity. Sci. Rep. 15, 6455 (2025). Author Contributions JS proposed experiments. SK conducted experiments. BSH supervised experiments, analyzed the data, and wrote the manuscript. Fund...

    2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.