{"id":"f37ca087-f93b-417a-9255-b0c4dcd0c851","arxiv_id":"1908.08595","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A meter-long, high-finesse fiber Fabry-Perot resonator interrogated by a diode laser locked to a twin reference cavity achieves strain resolutions down to 40 fε/√Hz at 23 kHz.","lead":"A meter-long fiber-optic cavity made of a single fiber with mirrored ends can detect strains as small as 40 femtostrain per root hertz at high frequencies, using an ordinary diode laser locked to a second identical fiber cavity. The key trick is to let the laser follow slow environmental drift by locking to a twin cavity kept in the same isolation box, keeping the sensor on its sharp resonance.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Twin-cavity drift cancellation is asserted but never quantified; without a residual differential frequency-noise measurement, the 1–100 Hz strain resolution is unsupported.","rationale":"Good-faith reading: the paper is an experimental demonstration; the central claim is a set of noise-limited strain resolutions. The key enabling assumption is that co-located twin cavities drift together enough to let a 10-Hz servo suppress sensor detuning drift, and that the PZT calibration is correct. The first is unsupported because no residual differential frequency noise is reported; the second is unsupported because the claimed independent verification is not shown. These are missing measurements, not internal contradictions. I do not see a red flag that warrants rejection: the high-frequency values are individually plausible, the cavity finesse and linewidth are characterized, and the side-detection scheme is standard. However, the absence of a residual-drift measurement means the 1–100 Hz number cannot be taken as robust, and the calibration gap affects all values. This is exactly the kind of missing support that the CONDITIONAL verdict should carry. Since the reader already reached CONDITIONAL for these reasons, my read does not change the verdict.","tokens_in":7941,"tokens_out":7434,"duration_ms":112457,"concrete_test":"While the laser is locked to FFPI-2 and no strain is applied to FFPI-1, record simultaneously the PDH error signals of both cavities (or the transmission fluctuations of FFPI-1) and compute the residual differential frequency-noise spectrum from 10 mHz to 100 kHz. Calibrate this spectrum to strain using the measured FFPI-1 resonance slope and an independently re-measured PZT scale (e.g., via a calibrated displacement actuator or an optical wavelength reference). If the calibrated residual noise exceeds the Fig. 5 floor in any band, or if the independent PZT scale changes the reported values by more than ~20%, the central resolution numbers need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reported 1–100 Hz resolution (800 fε/√Hz) is built on the twin-cavity common-mode rejection scheme described in Sec. II: 'Since both FFPIs experience the same environmental perturbations, their resonance peaks have similar drift.' The servo that locks the laser to the reference cavity has a bandwidth of only 10 Hz, so below ~10 Hz the laser tracks the reference resonance, and the sensor signal depends on the differential drift between the two cavities. The paper reports no measurement of that differential drift, no common-mode rejection ratio, and no residual relative frequency-noise spectrum for the two cavities. The 10–100 Hz portion of the same claim cannot rely on the reference at all; it depends on the absolute frequency stability of the sensor cavity, which is also not separately characterized. Because the low-frequency resolution is derived from the sensor noise floor rather than from calibrated strain injections at those frequencies, an uncancelled differential drift or absolute sensor jitter at the femtostrain level would directly mimic or mask the claimed floor. A secondary but related gap is the absolute strain scale: Sec. III states the PZT calibration 'has also been independently verified in our experiment,' but no verification data are shown, and that scale multiplies every reported fε/√Hz value.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports an experimental fiber-optic strain sensor based on a 1-m long, finesse-902 fiber Fabry-Perot interferometer (FFPI). The interrogation laser is a free-running diode laser frequency-locked to an identical reference FFPI, and the sensor is read out by side-of-fringe detection of transmitted power. Dynamic strain is applied through the sensor's PZT stretcher, calibrated using the manufacturer's specified response. The authors report a nominal strain resolution of 800 fε/√Hz from 1 to 100 Hz and values of 75, 60, and 40 fε/√Hz at 1 kHz, 2 kHz, and 23 kHz, respectively, and argue these are comparable to or better than phase-shifted and slow-light FBG sensors. The central technical claims are direct laser locking to a meter-long fiber cavity and common-mode environmental drift rejection between twin cavities.","tokens_in":8143,"tokens_out":6258,"duration_ms":57440,"significance":"If the reported resolutions are correct, the work is significant because it demonstrates that meter-scale high-finesse FFPIs can be used as ultrahigh-resolution strain sensors with inexpensive lasers, a combination not previously shown. The direct strain-response data at 300 Hz and the linearity curve in Fig. 4 are real experimental evidence, and the finesse is independently characterized. The main limitations are that the two pillars of the headline numbers—the twin-cavity common-mode drift rejection and the PZT strain calibration—are not quantitatively demonstrated in the manuscript. Since both are directly measurable, the central claims are plausible but need additional evidence before the femtostrain-level resolutions can be considered established.","major_comments":[{"comment":"The common-mode rejection of environmental drift between FFPI-1 and FFPI-2 is a load-bearing assumption but is never quantified. The servo bandwidth is stated as 10 Hz, so below roughly 10 Hz the laser tracks the reference resonance and the sensor signal depends on the differential drift of the two cavities; between 10 and 100 Hz it depends on the absolute stability of the sensor cavity. The manuscript provides no common-mode rejection ratio, no residual differential frequency-noise spectrum, and no long-term drift trace for the two cavities. Please add a direct measurement of the relative frequency fluctuations between the two cavities under the actual isolation conditions, and report the resulting contribution to the strain-equivalent noise over 1–100 Hz.","section":"Sec. II, Experimental Method"},{"comment":"The absolute scale of every reported femtostrain resolution is set by the PZT strain calibration, but the manuscript only states that the manufacturer-specified response 'has also been independently verified in our experiment' without showing the verification. Please provide the independent calibration data, for example a comparison of commanded PZT displacement against a known interferometric or mechanical reference, including the uncertainty in the calibration factor and the frequency range over which it applies.","section":"Sec. III, Experimental Result, strain calibration"},{"comment":"The low-frequency portion of the resolution spectrum is derived from the sensor noise floor, not from calibrated strain injections at those frequencies. The only calibrated dynamic-strain measurement shown is at 300 Hz (Fig. 4), so the conversion of the 1–100 Hz noise floor to strain units assumes the sensitivity is frequency-independent. Because the PZT transfer function and the cavity response can vary with frequency, please add calibrated strain injections at representative frequencies in the 1–100 Hz band, or justify the flat sensitivity assumption with a measured transfer function.","section":"Sec. III, Fig. 5"},{"comment":"No error bars or uncertainty budget are given for the measured strain-response curve or for the reported resolutions. Given that the headline claim is a set of specific numerical resolutions and a comparison with other techniques, the authors should state the statistical uncertainty, including the number of averages, resolution bandwidth, and calibration uncertainty, for each reported value.","section":"Sec. III, Figs. 4–5"}],"minor_comments":[{"comment":"The 'drift-over' condition is mentioned but not defined; please describe how it was established and how the free-running drift spectrum was recorded.","section":"Sec. II, Fig. 2(b)"},{"comment":"The abstract says 'within 1–100 Hz,' while the text says 'equal or below 800 fε/√Hz at any frequency above 2 Hz'; please reconcile the stated frequency range.","section":"Abstract and Sec. III"},{"comment":"The finesse is given as ~1000 in the introduction and as 902 in Sec. II; use the measured value consistently throughout.","section":"Introduction and Sec. II"},{"comment":"The spurious noise spikes are not all identified; labeling the 30-Hz spike and any others would help readers distinguish environmental artifacts from sensor noise.","section":"Sec. III, Fig. 5"},{"comment":"The attribution of the resolution limit to laser intensity and frequency noise is not backed by measurements; please add a relative intensity noise spectrum and a laser frequency-noise characterization, or soften the claim.","section":"Sec. IV, Conclusion"},{"comment":"The comparison with prior FFPI sensors (e.g., Gagliardi et al. [34]) and with FBG-based sensors would be more informative if the quoted competing resolutions were given with their measurement bandwidths and any stated uncertainties.","section":"Sec. III, comparison with prior work"}],"recommendation":"major_revision","confidential_remarks":"In my view the experiment is promising and the manuscript is close to a publishable letter, but the two quantitative anchors (common-mode rejection and PZT calibration) need to be demonstrated with data. If the authors can add a compact differential-drift measurement and an independent calibration check, perhaps as a supplement, the paper would meet the bar. I would not recommend rejection on the basis of the current concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Useful experimental letter. The genuinely new thing is that a 1-m fiber Fabry-Perot cavity with finesse ~900 can be interrogated with an off-the-shelf diode laser locked to an identical reference cavity, and that at kHz frequencies the measured strain resolution (75, 60, 40 fε/√Hz at 1, 2, 23 kHz) is plausibly at the level claimed. That is a real step beyond the 13-cm and 50-cm FFPI sensors in the prior literature, which relied on optical frequency combs or much more elaborate references.\n\nThe paper does several things right. It reports an independent finesse measurement (902) rather than trusting the spec sheet. It shows a linear strain-response curve and a 52 pε tone at 300 Hz with 32 dB SNR. The residual servo error spectra in Fig. 2 are consistent with the stated 10 Hz bandwidth. The high-frequency resolution numbers are not obviously inflated.\n\nThe soft spots are in the low-frequency part of the claim. The 800 fε/√Hz figure for 1–100 Hz leans on the statement that “both FFPIs experience the same environmental perturbations,” so locking to the reference cancels the sensor drift. That is an assertion, not a measurement. The text gives no common-mode rejection ratio, no residual differential frequency-noise spectrum between the two cavities, and no test of how well the two 1-m fibers stay in lockstep at 1–10 Hz. Since the servo bandwidth is only 10 Hz, below 10 Hz the sensor output is the difference of two slowly drifting cavities; above 10 Hz the reference loop is essentially not acting, so the sensor noise floor is whatever the isolation and the cavity do on their own. A femtostrain-level low-frequency resolution claim needs at least one of those measurements. The stress-test note is on target.\n\nAlso, the absolute strain scale comes from the manufacturer-specified PZT response; the paper says it was “independently verified” but shows no verification data. That factor multiplies every fε/√Hz number. It may be fine, but it needs to be shown or at least given an uncertainty budget. There are also no error bars anywhere.\n\nThe citation pattern is fine; self-citations are to the authors' own earlier work on thermal noise and cavity locking, which is relevant, and the comparison to phase-shifted/slow-light FBGs is standard in this community.\n\nBottom line: the high-frequency demonstration looks solid and useful; the low-frequency claim is under-supported. This deserves peer review—a good referee can pin down the missing common-mode and calibration measurements. I would not cite the 1–100 Hz number, but I would cite the kHz results as a proof of concept.","headline":"Solid high-frequency fiber Fabry-Perot sensing demonstration; the 1–100 Hz resolution claim leans on an unquantified common-mode rejection assumption and a manufacturer-only strain calibration.","tokens_in":8676,"tokens_out":2889,"would_cite":true,"duration_ms":37319,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A meter-long fiber Fabry-Perot cavity, interrogated by a diode laser locked to a twin reference cavity, achieves strain resolutions down to 40 femtostrain per root hertz, matching or beating phase-shifted and slow-light fiber Bragg…","keywords":["fiber Fabry-Perot interferometer","ultrahigh-resolution strain sensing","Pound-Drever-Hall locking","femtostrain resolution","fiber Bragg grating comparison","long-cavity sensing","dynamic strain measurement"],"falsifier":"Measure the relative frequency drift between the two cavity resonances over the 1-100 Hz band while both are isolated and the laser is locked; if the residual drift exceeds the equivalent of 800 femtostrain per root hertz, the low-frequency resolution claim fails. Independently calibrate the PZT-applied strain with a known reference, such as a strain gauge or a fiber Bragg grating, and check whether the reported 75, 60, and 40 femtostrain per root hertz values change.","tokens_in":7737,"feed_emoji":"🔬","tokens_out":9900,"duration_ms":291708,"temperature":0.7,"pith_summary":"The paper claims that a meter-long fiber Fabry-Perot resonator can serve as an ultrahigh-resolution strain sensor despite its large environmental drift, provided the probe laser is locked to a twin reference cavity exposed to the same perturbations. The authors report noise-limited resolutions of 800 femtostrain per root hertz in the 1-100 Hz band, improving to 75 femtostrain per root hertz at 1 kHz, 60 femtostrain per root hertz at 2 kHz, and 40 femtostrain per root hertz at 23 kHz, which they find comparable to or better than phase-shifted and slow-light fiber Bragg gratings. The practical significance is that these numbers come from an ordinary off-the-shelf diode laser and standard fiber components, with no frequency comb and no special laser stabilization. If the result holds, long-cavity FFPIs become a viable, simple route to femtostrain-level dynamic sensing.","feed_headline":"Meter-long fiber-optic sensor hits 40 fε/√Hz strain resolution","feed_subtitle":"Locking the probe laser to an identical cavity cancels drift, rivaling phase-shifted and slow-light gratings.","key_machinery":"The central object is a pair of nominally identical meter-long fiber Fabry-Perot interferometers (FFPIs): one acts as the sensor and the other as the reference. A Pound-Drever-Hall servo with a deliberately low 10-Hz bandwidth locks the interrogation diode laser to the reference cavity, so the laser tracks the slow, common-mode drift of both cavities while ignoring fast jitter. The sensor is read out by parking the laser on the steep side of a transmission resonance, converting strain-induced resonance shifts into transmitted-power changes. The high finesse (about 902) and long cavity (1 meter, 105-MHz free spectral range, 116-kHz linewidth) create narrow resonances that make this side-of-fringe discriminator highly sensitive.","core_discovery":"The paper reports the first ultrahigh-resolution strain sensor based on a meter-long fiber Fabry-Perot cavity with finesse around 900. By locking a diode laser to a second, identical FFPI placed in the same isolation enclosure, the laser follows the slow environmental drift of the cavity resonances, so the sensor resonance and the probe laser stay together for extended periods. Interrogating on the side of the transmission fringe, the authors measure a noise-limited strain resolution of about 800 femtostrain per root hertz from 1 to 100 Hz, improving to 75 femtostrain per root hertz at 1 kHz, 60 femtostrain per root hertz at 2 kHz, and 40 femtostrain per root hertz at 23 kHz. These numbers are comparable to or better than published results for phase-shifted and slow-light fiber Bragg gratings, and they are obtained with off-the-shelf components and no additional laser stabilization.","pith_inferences":["A direct measurement of the residual relative drift between the two cavities would yield the common-mode rejection ratio and should predict the 1-100 Hz noise floor; the paper does not report such a measurement.","If the sensor were read out with a Pound-Drever-Hall loop instead of side-of-fringe detection, laser intensity noise would be largely rejected, likely pushing the resolution toward the cavity's fundamental thermomechanical noise limit.","The manufacturer-specified PZT strain calibration sets the absolute scale of every reported femtostrain value; an independent calibration against a known strain reference would tighten confidence in the absolute numbers.","The scheme suggests a route to quasi-static or infrasonic strain sensing: with better thermal isolation and active drift compensation, the common-mode locking could hold for much longer timescales than demonstrated."],"forward_implications":["Meter-long FFPIs can serve as ultrahigh-resolution strain sensors at kilohertz frequencies, matching or exceeding the resolutions of phase-shifted and slow-light fiber Bragg gratings.","The twin-cavity locking scheme removes the need for optical frequency combs or specially stabilized narrow-linewidth lasers in long-cavity fiber sensing.","The low-frequency 1-100 Hz resolution is set by isolation quality and residual common-mode drift, so improving those should push the noise floor lower.","The same locking architecture can be transferred to other long fiber cavities for temperature, acoustic, or quasi-static sensing.","Side-of-fringe detection, despite its sensitivity to laser intensity noise, is sufficient for femtostrain-level dynamic measurements."],"supporting_citations":[{"why":"Reports the 140 femtostrain per root hertz at 1 kHz baseline for a frequency-locked phase-shifted FBG that the present sensor is compared against.","marker":"[12]"},{"why":"Reports the 30 femtostrain per root hertz at 30 kHz and 110 femtostrain per root hertz at 2 kHz slow-light FBG resolutions used as comparison benchmarks.","marker":"[18]"},{"why":"Demonstrates prior FFPI strain sensing at 220 femtostrain per root hertz at 1.5 kHz using an optical frequency comb, the main FFPI-based benchmark to surpass.","marker":"[34]"},{"why":"Reports a 50-cm FFPI strain sensor with 60 picostrain per root hertz near 900 Hz, the earlier long-cavity FFPI result this work extends.","marker":"[25]"},{"why":"Supplies the Pound-Drever-Hall frequency-locking method used to lock the diode laser to the reference cavity.","marker":"[37]"},{"why":"Documents the commercial fiber Fabry-Perot scanning interferometer used as both sensor and reference, including the specified finesse of 1000.","marker":"[36]"}],"fun_headline_variants":["Meter-long fiber cavity rivals slow-light gratings","Laser-locked fiber Fabry-Perot hits 40 fε/√Hz","Drift-canceling meter-long cavity for femtostrain sensing","High-finesse meter-long fiber rivals Bragg gratings","Meter-long FFPI achieves 40 fε/√Hz at 23 kHz"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result rests on the assumption that the two meter-long fiber cavities, sitting in the same insulated box, drift together so completely that locking the laser to the reference also cancels the sensor's slow drift; the paper does not report a direct measurement of how much of the sensor drift actually remains.","fun_headline_variants_meta":{"raw":{"variants":["Meter-long fiber cavity rivals slow-light gratings","Laser-locked fiber Fabry-Perot hits 40 fε/√Hz","Drift-canceling meter-long cavity for femtostrain sensing","High-finesse meter-long fiber rivals Bragg gratings","Meter-long FFPI achieves 40 fε/√Hz at 23 kHz"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000909,"raw_usage":{"total_tokens":3908,"prompt_tokens":949,"completion_tokens":2959,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":565,"completion_tokens_details":{"reasoning_tokens":2864}},"tokens_in":565,"tokens_out":2959,"duration_ms":22875,"temperature":1.0,"reasoning_tokens":2864,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:00:16.908026+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the relative frequency drift between the two cavity resonances over the 1-100 Hz band while both are isolated and the laser is locked; if the residual drift exceeds the equivalent of 800 femtostrain per root hertz, the low-frequency resolution claim fails. Independently calibrate the PZT-applied strain with a known reference, such as a strain gauge or a fiber Bragg grating, and check whether the reported 75, 60, and 40 femtostrain per root hertz values change.","supporting_citations":[{"cited_title":"Measuring attostrains in a slow-light fiber Bragg grating,","cited_arxiv_id":null,"evidence_quote":"Reports the 30 femtostrain per root hertz at 30 kHz and 110 femtostrain per root hertz at 2 kHz slow-light FBG resolutions used as comparison benchmarks."}],"review_version":1}