Montgomery Self-Imaging Without Lenses: FBA Handles for Revivals & Null Tests

Mar 5, 2026

•

C4 – Quantum Information & Channels (CPTP), C5 – Measurement & Open Systems (GKLS), C8 – Methods, Data & Reproducibility, Review EN

•

5 min read

The WinFuture article popularizes an optics result whose primary source is Yessenov et al. (arXiv 2025; Optica 2026): a direct observation of a spatially structured Montgomery effect (lensless self-imaging) that generalizes the Talbot effect to aperiodic structures and is claimed to show repeated strongly focused spots (~10 µm) in free space. According to the abstract, a dynamic optical hologram is used that discretizes radial spatial frequencies; self-imaging is reported for propagation distances of about 30 to 100 mm.

The FBA view translates this into an operational protocol: “preparation + hologram/propagation + camera” is treated as a controlled channel with an explicit measurement definition, “revival” as a measurable residual, and the whole claim is secured via hard pass/fail null tests against Fresnel/SLM artifacts.

Source anchors & subject

Submitted link

https://winfuture.de/news,156577.html#

Primary sources

https://arxiv.org/abs/2510.16269 (preprint; abstract mentions ~10 µm spots and 30–100 mm)
https://opg.optica.org/optica/fulltext.cfm?uri=optica-13-2-195 (Optica 13(2), 195–202 (2026))
https://opticapublishing.figshare.com/ndownloader/files/61291627 (dataset/supplement; exact contents open/unclear here)

Reality check

Standard/established: Free-space propagation in the paraxial regime is a linear (coherent) transformation; what is experimentally “seen” is always a specific detector output (pixel intensities after a measurement chain).
Standard/established: Talbot self-imaging is an established lensless recurrence phenomenon for periodic structures; “Montgomery” addresses (in the literature) self-imaging beyond strict periodicity.
Hypothesis: The primary works claim that a controlled discretization of radial spatial frequencies generates revivals/“self-imaging planes” (30–100 mm) with localized spots (~10 µm) and realizes recurring profiles including LG, HG, Ince–Gaussian, and Airy structures, among others.

FBA view

Handle: Model the setup as a controlled channel Φ(z,u): preparation → (propagation + hologram parameter u) → output state; “revival” then becomes a property of Φ at specific z, not an image narrative. (Definition III.4.1.1)
Principle: Camera readout as a POVM over pixels/ROIs; every “self-imaging” claim must be phrased as a claim about outcome distributions (including an error band). (Definition III.3.2.1; Formula box III.3.3.1)
Handle: Uncontrolled aberrations/partial coherence as an environment E (dilation); this clarifies which effects must be tested as “environment leakage” vs “signal.” (Formula box III.4.2.2)
Proxy: Define a revival residual R(z,u) as a normalized distance between images (or features) over the same ROI; revival planes are the z where R(z,u) shows minima, not where a screenshot looks pretty.
Principle: Composition logic: if the phenomenon is channel-structural, serially composed steps must be consistent (z scan as serial composition), rather than matching only pointwise. (Lemma III.4.2.1)
Control idea: “Independent control of spot size and revival period” as a separability test in parameter space: local changes u_spot vs u_period should steer observables (R, spot metrics) in distinct, preregistered directions, rather than renaming the same trade-off.

New insights from FBA

FROM→TO: “Self-imaging” as picture similarity → revival as a residual minimum R(z,u) under a fixed POVM/ROI. Implicit assumption: the chosen measurement definition (ROI, normalization, drift correction) is stable and documented.
FROM→TO: “3D light patterns” as a demo → plane indexing as a test object: cross-talk between planes becomes measurable as a residual matrix over z targets. Implicit assumption: z scanning does not systematically change the process (exposure, alignment, focus drift).
FROM→TO: “Discretizing radial spatial frequencies” as a story → leakage test: revival strength becomes a function of spectral leakage (sidebands) and is thus experimentally falsifiable. Implicit assumption: u truly parametrizes the relevant spectral discretization rather than merely acting as an effective lens.
FROM→TO: “Mode revivals” as a qualitative gallery → mode fidelity as a measurable quantity (e.g., overlap/classifier) with a fail criterion. Implicit assumption: mode decomposition is calibrated and robust to aberrations.

Clarification / improvement with FBA

Confounder: Partial coherence (source/spectrum) can damp or shift revivals. Control idea: log coherence proxies (spectral width, effective source size) and evaluate R(z,u) in coherence bins.
Confounder: The SLM pixel lattice can inject Talbot-like scales. Control idea: order filtering/aperture scans and deliberate changes of effective sampling scale (resampling/mask changes) at constant u.
Control idea: “Hologram shuffle”: same spectral envelope, but random assignment/phases within the radial bins; expectation: genuine Montgomery revival structure collapses when the bin structure is destroyed.
Proxy: Minimal reproducibility package: raw images for the z scan, parameters u, a clear definition of ROI/normalization/drift correction, plus code to compute R(z,u); this makes the claim independent of single showcase images.

Alternative readings & conclusions

Hypothesis: Part of the observed “refocusing” could be explained by an effective quadratic phase (effective lens) or aperture focusing; then the revival signal is more focus wander than self-imaging.
Hypothesis: Recurrence scales could be driven by hardware periodicities (SLM pitch/detector sampling); then the relevant z scales should co-scale with pixel size or with filter settings.
open/unclear: “Independent control” (spot size vs revival period) may hold only in a narrow regime; without an explicit, broad parameter space, generality remains unclear.

Tests/Experiments (Pass/Fail) with an FBA touch

Null test (Standard/established): Talbot benchmark | periodic grating, same optics/camera | Talbot distances/patterns appear consistently | deviation beyond error band/calibration
Residual (FBA): R(z,u) minima | z scan (30–100 mm) with varying u | clear, reproducible residual minima + u-dependent shift | no reproducible minima or u has no effect
Null test (Hypothesis): hologram shuffle | same envelope, but bin phases/assignment randomized | revivals collapse (R minima disappear) | revivals remain stable (artifact suspicion)
Pass/Fail (open/unclear): sampling scaling | vary effective pixel/lattice scale (resampling/mask changes) | Montgomery scales remain (nearly) invariant | z scales co-scale with hardware periodicity

Added value of the FBA view

Added value: 8/10 – The post shifts from a visual demonstration to a clear channel/measurement definition plus a residual and null-test suite that separates genuine revivals from hardware and coherence artifacts.