{"id":"b39aac58-7cea-4a00-a81b-16c970e85f21","arxiv_id":"2509.10235","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Off-axis digital holography is shown to image nanometre-scale fluctuations of superfluid helium free surfaces and to recover the gravity-capillary dispersion relation in two types of cryostat.","lead":"This paper demonstrates off-axis digital holography as a full-field, nanometre-sensitive way to image the free surface of superfluid helium in both a bath cryostat and a cryogen-free refrigerator. It reconstructs the dispersion relation of gravity-capillary surface waves from noise-excited normal modes, opening a minimally invasive readout for quantum fluid experiments.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Dry-cryostat dispersion benchmark is self-calibrating: a fitted magnification rescales all measured wavenumbers while h0 is fitted to the same curve, so agreement with Eq. (2) may not independently confirm the wavenumber calibration.","rationale":"The central claim is that off-axis digital holography provides a quantitative, full-field readout of superfluid helium surfaces, benchmarked by recovering the gravity-capillary dispersion relation. The wet-cryostat result supports this claim strongly, because the measured frequencies and fitted wavenumbers fall on Eq. (2) without fitted physics. The dry-cryostat result is the weakest link: the agreement with Eq. (2) is achieved by fitting both h0 and a global magnification that directly rescales the measured wavenumbers. This is not circular in the strict sense, but it substantially weakens the benchmark, because the fit can absorb a systematic error in the spatial calibration. The paper itself provides a qualitative explanation for the lensing (Fig. S8) and shows consistency across azimuthal orders in Fig. S7, which is nontrivial, but all those fits share the same constant-magnification assumption. An independent calibration of the magnification would settle whether the dispersion collapse is physical or an artifact of the two-parameter fit. This concern is addressable and does not invalidate the imaging demonstration, so the CONDITIONAL verdict remains appropriate. Data availability is restricted to 'upon reasonable request', which limits independent verification but is not the most load-bearing scientific concern.","tokens_in":17489,"tokens_out":8035,"duration_ms":71595,"concrete_test":"Independently calibrate the pixel-to-physical-length scale at the sample plane in the cryogen-free setup (e.g., by imaging a known calibration reticle placed at the bottom optical port under the same cryogenic conditions, including the superfluid film), then re-fit the m = ±2 dispersion data with the magnification fixed and h0 as the only free parameter. If the fixed-magnification fit no longer collapses onto Eq. (2) (residuals increase systematically) or yields h0 outside the 578 ± 21 µm interval, the quantitative benchmark in Sec. II C fails. As a secondary check, fit the magnification separately for modes with low and high radial index n; a statistically significant n-dependence would invalidate the constant-magnification model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Sec. II C (Fig. 4e), the cryogen-free experiment validates the dispersion relation by fitting Eq. (2) with two free parameters: the film thickness h0 and a global prefactor ('magnification') that rescales all measured wavenumbers, introduced to correct for the superfluid meniscus acting as a plano-concave lens (Fig. S8). The wet-cryostat result (Fig. 2d) is a stronger validation because it uses no fitted physics; the dry-system benchmark does not independently confirm the height calibration or the lensing model. Because a two-parameter fit of a monotonic dispersion curve can accommodate a global k-rescaling and a thickness parameter, the good fit does not by itself prove that the reconstructed wavenumbers (and hence the heights) are correct. The specific risk is that the fitted magnification (1.174 ± 0.004) absorbs systematic errors from non-constant meniscus lensing or from an incorrect radial coordinate mapping; if the magnification is actually radius-dependent, modes with different radial index n would be rescaled differently, and the inferred h0 = 578 ± 21 µm would be biased, as would the shallow-water speed c = 75 ± 2 mm/s quoted in the Discussion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the first implementation, to the authors' knowledge, of off-axis digital holography (DH) as a full-field optical readout for the free surface of superfluid 4He, demonstrated in two cryogenic platforms: a helium bath cryostat and a cryogen-free refrigerator. The method reconstructs time-resolved surface height maps from holographic phase shifts, resolves individual normal modes via spectral filtering and principal-component analysis, and recovers the gravity-capillary dispersion relation, Eq. (2). In the wet cryostat, mode frequencies and Bessel-function spatial profiles are compared with predictions based on a Neumann boundary condition and no fitted physics; in the dry system, the dispersion is recovered by fitting Eq. (2) with two free parameters, a film thickness h0 and a global wavenumber magnification factor, yielding h0 = (578 ± 21) µm.","tokens_in":17715,"tokens_out":5334,"duration_ms":50666,"significance":"If the claims hold, this is a valuable new capability for studying superfluid surface dynamics, with clear relevance to wave turbulence, analogue gravity, and quantum optomechanics. The wet-cryostat results provide strong internal support: the reconstructed spatial profiles of individual modes match Bessel functions, the azimuthal decomposition is consistent with m = ±1 modes, and the dispersion points follow Eq. (2) without fitted physics in the wet system. The dry-system demonstration is a useful proof of concept but is weaker as a quantitative benchmark because the magnification and thickness are fitted to the same dispersion relation. The paper does not provide public data or code, limiting independent verification of the quantitative claims.","major_comments":[{"comment":"The cryogen-free dispersion benchmark is self-calibrating and therefore does not independently validate the wavenumber reconstruction. The two fitted parameters—the film thickness h0 and a global prefactor that rescales all measured wavenumbers—both determine the theoretical curve, so agreement with Eq. (2) can absorb systematic errors in the radial-coordinate mapping. If the meniscus acts as a radius-dependent lens rather than a constant magnification, modes with different radial index n would be rescaled differently, biasing the inferred h0 = (578 ± 21) µm and the shallow-water speed c = (75 ± 2) mm/s quoted in the Discussion. The authors should provide an independent calibration of the magnification (e.g., imaging a known test pattern through the same film or measuring h0 by an independent technique) or explicitly reframe the dry-system result as an internal consistency check rather than an independent validation of the reconstructed wavenumbers.","section":"§II C, Fig. 4e, Eq. (2)"},{"comment":"The identification of modes A–F is partly circular. In Fig. 2a the spectral peaks are labeled by matching their frequencies to values predicted from Eq. (2) with the assumed Neumann boundary condition; these labels are then used to organize the dispersion points in Fig. 2d. Although the subsequent Bessel-profile fitting provides independent wavenumber information, the radial fits operate within a small central field of view (≈6.3 mm radius) where different radial indices n are visually similar, as the text itself states. The authors should quantify how sensitive the fitted wavenumbers are to the assumed (m, n) assignment, for example by allowing the mode indices to be selected purely from the spatial data or by reporting fit residuals for alternative assignments.","section":"§II B, Fig. 2a and Fig. 2d, Eqs. (1)–(2)"},{"comment":"The wet-cryostat mode labels and the resulting dispersion relation assume a Neumann (free-slip) boundary condition at the cell wall. If the actual contact line is pinned (Dirichlet) or partially pinned, all predicted mode frequencies shift and the assignment of observed peaks to (m, n) changes. The agreement shown in Fig. 2a is presented as preliminary justification, but a match between observed peaks and one boundary-condition choice is not an independent test of that choice. Please report the sensitivity of the recovered dispersion to the boundary condition, for example by comparing the predicted frequencies for Dirichlet conditions and stating whether the same modes A–F would remain the nearest assignments to the observed peaks.","section":"§II B, Eq. (2) and boundary condition at R = 27 mm"}],"minor_comments":[{"comment":"The phrase 'machine-learning-based analysis' is used for PCA and truncated SVD, which are classical linear dimensionality-reduction techniques; consider describing them more precisely as 'statistical signal-decomposition methods' to avoid overstating the machine-learning content.","section":"Abstract and §II B"},{"comment":"The vertical lines marking predicted Bessel-mode frequencies in Fig. 2a are not accompanied by a stated scaling or normalization; please specify how their heights are chosen so that the visual alignment is interpretable.","section":"§II B, Fig. 2a inset"},{"comment":"The histograms in Fig. S3 are normalized by image size in pixels, but the text reports an accuracy 'better than 2 mm'; please provide the pixel-to-millimetre conversion in the caption.","section":"Fig. S3 caption"},{"comment":"The sentence 'The latter parameter was introduced to adjust the divergence created by the superfluid sample acting as a plano-concave lens' is ambiguous; please specify that the divergence is in the probe beam and explain how a single constant magnification factor accounts for a spatially varying lens effect.","section":"§II C"},{"comment":"The value c = (456 ± 22) mm/s for the wet system is quoted without explaining whether it comes from a linear fit to the lowest-frequency data points or from the relation c = sqrt(gh0) using h0 = 20 mm; please add a sentence clarifying the derivation.","section":"Discussion"}],"recommendation":"major_revision","confidential_remarks":"The central proof-of-concept is likely sound, especially for the wet-cryostat system, but the dry-system benchmark needs an independent calibration or a clear reframing as a consistency check rather than a validation. I would also encourage the editor to ask the authors to make the processed data and analysis code available, since the quantitative claims (h0, c, mode assignments) are central and the current 'available upon request' policy makes independent verification difficult. The manuscript fits the scope of physics.optics."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe headline: the wet-cryostat half of this paper is the real advance; the dry-cryostat dispersion \"benchmark\" is partially self-validating, so don't cite it as independent confirmation.\n\nWhat's actually new: this is the first adaptation of off-axis digital holography to superfluid helium surfaces, done in two cryogenic environments. That's a useful capability for quantum fluids and analogue gravity. The paper does a couple of things well. The mode decomposition in the bath cryostat is careful: they identify spatial Bessel profiles, fit the symmetry origin, and recover a dispersion relation that matches Eq. (2) without any fitted physics. That is a strong internal check. The evaporation-rate monitoring is a nice incidental result.\n\nThe soft spots are concentrated in the cryogen-free section. The fit to Eq. (2) uses two free parameters—film thickness h0 and a global magnification prefactor—and both are fitted to the same dispersion curve. The magnification is supposed to correct for meniscus lensing, but if that lensing is radius-dependent, a single constant cannot capture it, and the fitted h0 (578±21 µm) and shallow-water speed (75±2 mm/s) inherit any bias. The agreement in Fig. 4e is therefore suggestive but not a closed loop. The supplementary fits for other azimuthal modes are consistent, which helps, but they use the same fitting protocol. Also, the analysis selects only modes that fit well; that's normal, but it should be acknowledged as selection. And 'data available upon request' is weak for a paper whose central claim depends on quantitative phase reconstruction.\n\nThe stress-test note you passed me is right about the dry-system circularity, but the wet-cryostat result stands on its own. The paper is honest about its limitations; no red flags in the citation pattern; the machine-learning language is mostly PCA/SVD, which is fine but could be described more plainly.\n\nWho is this for? Experimentalists in superfluid hydrodynamics and analogue gravity who need a full-field surface probe. The method paper deserves a serious referee: the claims are concrete, the wet benchmark is solid, and the dry-system weaknesses are addressable in revision. I'd accept it for review with a request for a more transparent accounting of the dry-system fitting, plus raw data or processed data release for at least one dataset.","headline":"Wet-cryostat DH demonstration is a genuine advance; the dry-cryostat dispersion fit partly validates itself, so treat that section as suggestive rather than confirmatory.","tokens_in":18351,"tokens_out":1789,"would_cite":true,"duration_ms":16224,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Digital holography captures the full-field surface shape of superfluid helium for the first time.","keywords":["digital holography","superfluid helium","free-surface waves","gravity-capillary waves","cryogenic optics","normal-mode analysis","surface topography","thin helium films"],"falsifier":"Place a calibration grid in the sample cell and record holograms with and without the superfluid film; if the required wavenumber rescaling is not a single radius-independent constant but varies across the field of view, the global-magnification model and the inferred film thickness are falsified. Independently, comparing the fitted film thickness with a direct interferometric thickness measurement would settle the dispersion-based benchmark.","tokens_in":1904,"feed_emoji":"🌊","tokens_out":2300,"duration_ms":435896,"temperature":0.7,"pith_summary":"Superfluid helium's free surface is notoriously hard to image: its refractive index is barely different from vacuum, cryogenic enclosures restrict optical access, and cooler vibrations shake the optics. This paper reports that off-axis digital holography can overcome those obstacles, reconstructing full-field surface height maps of superfluid 4He at nanometre-to-micrometre amplitudes in both a helium bath cryostat and a cryogen-free refrigerator. From the reconstructed height fields, the authors isolate noise-excited normal modes and show that their frequencies and wavenumbers follow the gravity-capillary dispersion relation. In the cryogen-free system, the dispersion fit also yields the thickness of a sub-millimetre superfluid film, a quantitative benchmark for the method. If the claims hold, digital holography becomes a minimally invasive, broadly deployable readout for superfluid surface dynamics, from thin films to quantum turbulence.","feed_headline":"Digital holography captures superfluid helium surface waves","feed_subtitle":"Nanometre-scale ripples in two cryostats match the gravity-capillary dispersion relation.","key_machinery":"The load-bearing object is the off-axis digital hologram: an interference pattern between a reference beam and a probe beam that has passed through or reflected from the superfluid, recorded by a high-speed camera. A spatial Fourier filter isolates the +1 diffraction order, and phase differences between frames are converted to surface height using the refractive-index contrast $\\Delta n\\simeq0.027$, with the factor-of-two difference between reflective and transmissive geometries. Identification of the observed waves is carried by the classical normal-mode ansatz, Bessel functions in a cylinder with free-slip walls, and by the gravity-capillary dispersion relation, which turns resolved mode frequencies and wavenumbers into a quantitative test. Machine-learning decomposition, through principal component analysis or truncated singular value decomposition, separates noise-driven modes from the background and locates the cell symmetry centre, while in the dry system a fitted global magnification corrects for the meniscus acting as a plano-concave lens.","core_discovery":"The central claim is that off-axis digital holography works as a quantitative full-field readout for the free surface of superfluid 4He, despite helium's low refractive index contrast and the mechanical noise of cryogenic systems. The probe beam's phase shift after traversing or reflecting from the sample is converted to surface height through $\\phi/(2\\pi)=\\Delta n\\,h/\\lambda$ in transmission, with a factor of two larger shift in reflection, and comparing each hologram with a reference frame removes static aberrations. In a 54-mm cylindrical bath-cryostat cell, six low-frequency modes are matched to Bessel normal modes $J_{|m|}(k_{mn}r)e^{im\\phi}$, and up to sixteen counter- and co-rotating one-fold pairs are resolved; their frequencies and wavenumbers agree with $\\omega_{mn}^2=(g+\\sigma k_{mn}^2/\\rho)k_{mn}\\tanh(h_0 k_{mn})$. In a cryogen-free refrigerator, a one-fold $(m,n)=(1,2)$ mode locates the cell centre, and $m=\\pm2$ modes up to 90 Hz follow the same relation when a single global magnification of $1.174\\pm0.004$ and a film thickness $h_0=(578\\pm21)\\,\\mu$m are fitted. The paper presents this as a proof of principle that digital holography can be integrated into traditional and cryogen-free platforms.","pith_inferences":["A natural next test is to push the same readout to nanometre-thin superfluid films, where third-sound waves and substrate interactions dominate; the reported nanometre sensitivity makes this plausible, but the paper does not demonstrate it.","If the single-magnification lensing model holds, the same holographic data could be inverted locally to map film-thickness variations across the field of view, effectively turning the meniscus distortion into a diagnostic.","Reducing mechanical noise should bring the technique close to the predicted roughly 7-nm dimples of quantised vortices, since the phase sensitivity already resolves sub-micrometre waves; this is an extrapolation, not a result of the paper.","The measured shallow-water speed of $(75\\pm2)$ mm/s in the film suggests the method could serve analogue-gravity experiments seeking non-dispersive surface waves, though the paper only notes the platform's potential."],"forward_implications":["Digital holography gives quantitative, full-field surface topography of superfluid helium, not just local slope measurements, over centimetre-scale fields of view.","In a cylindrical cell, ambient mechanical noise alone excites enough normal modes to reconstruct the gravity-capillary dispersion relation without external wave driving.","For thick superfluid films, the dispersion-relation fit returns the absolute film thickness, turning the readout into a self-calibrating thickness measurement.","The method works in cryogen-free refrigerators as well as helium-bath cryostats, so it can be added to existing low-temperature platforms without major optical redesign.","The global phase drift also tracks helium evaporation in real time, giving a continuous measure of mass loss alongside wave imaging."],"supporting_citations":[{"why":"Establishes the digital-holography surface-profilometry method and its nanometre amplitude sensitivity, which this work adapts to superfluid helium.","marker":"[20]"},{"why":"Supplies the synthetic Schlieren full-field topography method that digital holography extends and contrasts with for superfluid surfaces.","marker":"[11]"},{"why":"Demonstrates synthetic Schlieren imaging of superfluid helium surfaces, providing the prior state of the art for full-field surface readout.","marker":"[14]"},{"why":"Supplies the helium properties, including viscosity, surface tension, and density, that enter the dispersion relation and damping estimates.","marker":"[29]"},{"why":"Gives the gravity-capillary dispersion relation for linear surface waves used to test the reconstructed modes.","marker":"[30]"},{"why":"Provides the randomized principal component analysis algorithm used to isolate noise-driven normal modes.","marker":"[31]"},{"why":"Supplies the reference-conjugated hologram method used to subtract static aberrations in phase retrieval.","marker":"[28]"},{"why":"Provides the spatial filtering procedure for zero-order and twin-image elimination in off-axis digital holography.","marker":"[48]"}],"fun_headline_variants":["Holography reveals superfluid helium surface ripples","Digital holography reads superfluid helium waves","Holographic imaging maps superfluid helium surfaces","Nanoscale ripples on superfluid helium caught by holography","Holography tracks superfluid helium surface waves in two cryostats"],"cache_read_input_tokens":20352,"weakest_assumption_plain":"The quantitative dispersion and thickness results rest on two modelling assumptions: the helium meniscus distorts the image by a single constant magnification, and the surface slips freely at the cell wall; if either is wrong, the extracted wavenumbers and the fitted film thickness are biased.","fun_headline_variants_meta":{"raw":{"variants":["Holography reveals superfluid helium surface ripples","Digital holography reads superfluid helium waves","Holographic imaging maps superfluid helium surfaces","Nanoscale ripples on superfluid helium caught by holography","Holography tracks superfluid helium surface waves in two cryostats"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000854,"raw_usage":{"total_tokens":3769,"prompt_tokens":1062,"completion_tokens":2707,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":678,"completion_tokens_details":{"reasoning_tokens":2624}},"tokens_in":678,"tokens_out":2707,"duration_ms":16817,"temperature":1.0,"reasoning_tokens":2624,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:57:15.120808+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Place a calibration grid in the sample cell and record holograms with and without the superfluid film; if the required wavenumber rescaling is not a single radius-independent constant but varies across the field of view, the global-magnification model and the inferred film thickness are falsified. Independently, comparing the fitted film thickness with a direct interferometric thickness measurement would settle the dispersion-based benchmark.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the digital-holography surface-profilometry method and its nanometre amplitude sensitivity, which this work adapts to superfluid helium."},{"cited_title":"Moisy, M","cited_arxiv_id":null,"evidence_quote":"Supplies the synthetic Schlieren full-field topography method that digital holography extends and contrasts with for superfluid surfaces."},{"cited_title":"ˇSvanˇ cara, P","cited_arxiv_id":null,"evidence_quote":"Demonstrates synthetic Schlieren imaging of superfluid helium surfaces, providing the prior state of the art for full-field surface readout."},{"cited_title":"Colomb, J","cited_arxiv_id":null,"evidence_quote":"Supplies the helium properties, including viscosity, surface tension, and density, that enter the dispersion relation and damping estimates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the gravity-capillary dispersion relation for linear surface waves used to test the reconstructed modes."},{"cited_title":"Whitham, Linear dispersive waves, inLinear and Non- linear Waves(John Wiley & Sons, 1999) Chap","cited_arxiv_id":null,"evidence_quote":"Provides the randomized principal component analysis algorithm used to isolate noise-driven normal modes."},{"cited_title":"Pobell,Matter and Methods at Low Temperatures, 3rd ed","cited_arxiv_id":null,"evidence_quote":"Supplies the reference-conjugated hologram method used to subtract static aberrations in phase retrieval."},{"cited_title":"Harris, On the use of windows for harmonic analy- sis with the discrete Fourier transform, Proc","cited_arxiv_id":null,"evidence_quote":"Provides the spatial filtering procedure for zero-order and twin-image elimination in off-axis digital holography."}],"review_version":2}