{"id":"4c1b5917-bd36-4440-b667-8254eece1ef0","arxiv_id":"2608.04372","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A commissioned photonic lantern spectrograph on Subaru/SCExAO reaches 35-50 microarcsecond spectro-astrometric precision, recovering stellar spin axes and disk structure below the diffraction limit.","lead":"FIRST-PL, a new visible-light spectrograph on Subaru, uses a photonic lantern to split starlight into 19 fibers and measure positions of spectral features far below the telescope's diffraction limit. It reports 40% injection efficiency and microarcsecond-level spectro-astrometry on bright stars, potentially replacing interferometer measurements for some targets.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No on-sky null test or response-matrix error budget is presented; the 35-50 µas spectro-astrometry claim rests on a calibration-stability assumption that §5.1 itself shows was fragile.","rationale":"The reader's weakest_assumption is exactly where I focus: the calibration of the PL spatial response. The strongest independent evidence for the central claim is the Altair spin-axis PA agreeing with long-baseline interferometry; that is real support and justifies a conditional rather than reject posture. However, the manuscript's own §5.1 shows the passive-jitter calibration path had to be changed because AO residuals shrank, which undermines the claim that §3.1's passive diversity is generally sufficient. Without a null test or a calibrator-calibrator repeatability measurement, the claimed 35-50 µas precision cannot be separated from response-matrix drift. I therefore agree with the reader and recommend keeping the verdict CONDITIONAL, pending the companion-paper validation or the null test above. I do not find an internal inconsistency in the throughput or sensitivity estimates that would change the verdict; the unsupported 12× SMF improvement is secondary and not load-bearing.","tokens_in":12305,"tokens_out":5741,"duration_ms":61893,"concrete_test":"Run the identical reduction on a bright unresolved star observed immediately before or after Altair with the same TT/static-jitter protocol, and require the extracted photocenter versus wavelength to be consistent with zero at <10 µas across the Hα line. If the null-star photocenter exceeds 35 µas, or if using the simultaneous VAMPIRES PSF centroid time series as a regressor shifts the Altair spin-axis PA by more than 1σ, the headline precision is dominated by response-calibration systematics rather than astrophysical signal.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The manuscript's central claim is that the 19-port PL intensity response (wavelength-resolved across Hα) can be inverted to a 35-50 µas photocenter with the residual attributed to astrophysics. The load-bearing condition is that the PL response matrix is stable and linear between calibrator and science, and that residual tip-tilt jitter is sufficient to map it. §3.1 asserts jitter provides passive spatial diversity; §4.3 invokes simultaneous VAMPIRES/NIR PSF telemetry; but the paper gives no on-sky null test on an unresolved star, no calibrator repeatability measurement, and no error budget separating photon noise from response-model mismatch. Any slow drift in non-common-path aberrations between the 10% VAMPIRES path and the 90% FIRST-PL path, differential flexure, or PL mode-coupling changes will produce a wavelength-dependent pseudo-photocenter that can mimic the 0.27 mas minor-axis asymmetry and bias the spin-axis PA. The Altair PA agreement with VLTI/CHARA is encouraging, but with a claimed uncertainty of only ±2.8°, a calibration systematics of a few degrees is not excluded by that single comparison. §5.1 concedes that improved AO reduced the passive jitter diversity and the authors had to introduce manual PSF dithering and later active TT modulation—an implicit admission that the passive calibration basis was not robust. The companion papers (Kim et al. 2025; Walk et al. in prep.) may contain the needed validation, but it is not in this manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the commissioning of FIRST-PL, a 19-port visible photonic lantern spectrograph on Subaru/SCExAO. It describes the instrument architecture, three observing modes (spectro-astrometry, on-axis and off-axis image reconstruction), on-sky throughput performance, and early science demonstrations: β CMi Hα spectro-astrometry with 50 µas precision and a 0.27 mas minor-axis asymmetry, Altair spin-axis PA -61.8°±2.8° from 35 µas precision measurements, and a 5σ raw contrast limit of ~10^-3 at 100 mas for HIP 81126. The paper claims a 12× throughput improvement over single-mode fibers and announces open community access starting in Semester 2026B. Detailed data reduction and calibration theory are deferred to companion papers (Kim et al. 2025; Walk et al. in prep.; Sarrazin et al. in prep.).","tokens_in":12593,"tokens_out":3844,"duration_ms":35557,"significance":"If the spectro-astrometric precision figures hold up, FIRST-PL demonstrates that photonic-lantern-based single-aperture spectro-astrometry can reach tens of microarcseconds on bright stars, competing with long-baseline interferometry for selected measurements such as stellar spin axes and disk kinematics. The external consistency of the Altair position angle with VLTI/CHARA within 1σ provides an independent anchor for the method and is a genuine strength. The paper also documents a technically interesting instrument with real on-sky commissioning results, including injection efficiency measurements and a first contrast curve. However, because the calibration pipeline, null tests, and systematic error budget are not presented here, the quantitative science claims must be treated as preliminary until the companion papers provide the missing validation.","major_comments":[{"comment":"The central claim of 35–50 µas spectro-astrometry rests on the assumption that the PL response matrix is stable and linear between calibrator and science target, and that residual tip-tilt jitter provides sufficient spatial diversity to map it. The manuscript provides no on-sky null test on an unresolved star, no repeated calibrator measurements, and no error budget separating photon noise from response-model mismatch. The only external validation is the Altair PA agreement with interferometry; with a claimed uncertainty of ±2.8°, a few degrees of calibration systematics are not excluded by that single comparison. Please provide a null test, calibrator repeatability, and an explicit error budget, or clearly mark the science results as preliminary and dependent on the companion papers.","section":"§4.3 and §3.1"},{"comment":"Section 5.1 states that as AO performance improves, natural tip-tilt jitter decreases significantly, complicating spectro-astrometric calibration and requiring manual PSF dithering, and that a dedicated TT-modulation mode was recently implemented and validated on-sky. This directly undercuts the §3.1 operational principle that passive jitter provides 'inherent motion' sufficient to map the PL response without active modulation. If the static-mode calibration basis was fragile, the static data used for the β CMi and Altair results need to be re-analyzed with the new modulated mode to demonstrate that those results are robust, or the paper should explicitly state the limitation.","section":"§5.1"},{"comment":"The abstract claims a '12× throughput improvement over single-mode fibers,' but §4.1 reports only on-sky injection efficiency (21% and 40% mean at 642 nm and 680 nm) and an end-to-end throughput of ~1%; no direct comparison measurement against a single-mode fiber under the same conditions is shown. Please provide the measurement basis for the factor of 12, or remove the claim.","section":"§4.1 and Abstract"},{"comment":"The sensitivity and astrometric precision statements are not tied to a quantitative error budget. Equation (1) gives detector noise per pixel, and the text asserts σ_astro ∝ 1/SNR, but there is no derivation of how photon noise propagates through the PL modal response inversion to the final photocenter precision, nor any accounting for systematic terms (e.g., differential flexure or non-common-path aberrations between the 10% VAMPIRES path and the 90% FIRST-PL path). The reported 35–50 µas values therefore lack a demonstrated statistical foundation in this manuscript.","section":"§4.2"}],"minor_comments":[{"comment":"There is a typo in the first sentence: 'Engineering bservations' should read 'Engineering observations.'","section":"§4.1"},{"comment":"The instrument name 'V AMPIRES' is written with an internal space; elsewhere and in the references it appears as 'VAMPIRES'. Please make the spelling consistent.","section":"§3.1"},{"comment":"The target Humu is introduced as 'Humu (Altair, α Aql)' but later abbreviated inconsistently as 'Humu'; please define the preferred name at first use and use it consistently.","section":"§4.3.2"},{"comment":"The target is written as 'β-CMi' in the abstract and 'βCMi' in the text; please use a single notation throughout.","section":"Abstract and §4.3.1"},{"comment":"The field-of-view entry '80 mas @ f/8' does not specify the wavelength at which this was measured; the FoV of a photonic lantern is wavelength dependent and should be qualified.","section":"Table 1"},{"comment":"References [29] and [34] are SPIE proceedings with 2026 dates; please verify that they are publicly available or provide DOIs, since the manuscript relies on them for key claims.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is honest about its dependence on companion papers, but as written it does not provide sufficient internal support for the headline science numbers. The Altair external validation is a genuine strength. The editor may consider whether a commissioning paper can carry the full science claims without the pipeline papers, or whether the quantitative claims should be trimmed to instrument performance only. The 12× throughput claim appears unsubstantiated and should be corrected or removed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"FIRST-PL is a credible commissioning report, but if you're hoping for a new science result, the headline spectro-astrometry numbers are re-reported from Kim et al. (2025) and Walk et al. (2026). What is genuinely new is the instrument work: the piezo tip-tilt modulator, the calibration unit, the Wollaston polarimetric mode, on-sky injection efficiencies at 642 and 680 nm, and the HIP 81126 contrast curve. That part is solid and useful.\n\nThe strongest evidence in the paper is the Altair spin-axis position angle, which agrees with VLTI/CHARA within 1σ. That is an independent anchor for the method and should not be discounted. The Beta CMi repeated epochs also suggest repeatability.\n\nThe soft spots are the ones the stress-test note names. The 12x throughput improvement over SMF is asserted without a measurement in the text; you get injection efficiencies and end-to-end ~1% throughput, but no direct comparison. More importantly, the 35-50 µas precision claims rest on the stability and linearity of the PL response matrix, and the paper gives no null test on an unresolved star, no calibrator repeatability, and no error budget separating photon noise from response-model mismatch. Section 5.1 is an implicit admission that the passive jitter diversity was not enough: they had to introduce manual PSF dithering and then active TT modulation. That doesn't invalidate the Altair result—it was externally validated—but it does mean the precision numbers as quoted in this paper cannot be independently checked. The 0.27 mas disk asymmetry is at a level where response-matrix systematics could mimic the signal. This is a real concern, not manufactured.\n\nI would send this to a serious referee, but the referee should insist that either the companion papers' reduction details are included or a null test / response-matrix error budget is added. As a standalone paper, it is a useful status report for the instrument and a reasonable citation for the hardware.","headline":"A credible commissioning report whose real advance is the hardware; the headline science numbers are re-reported and their calibration basis is deferred to companion papers, so referee it but demand the missing error budget.","tokens_in":13230,"tokens_out":2511,"would_cite":true,"duration_ms":24806,"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 photonic lantern spectrograph on Subaru/SCExAO reaches 35 microarcsecond astrometry, recovering Altair's spin axis and a beta CMi disk asymmetry.","keywords":["photonic lantern","spectro-astrometry","sub-diffraction-limit astronomy","Subaru/SCExAO","visible spectroscopy","high-contrast imaging","integral field spectroscopy","stellar spin axes"],"falsifier":"A decisive check is to feed an unresolved calibration source through the same optics and tip-tilt dither pattern used on sky; if the pipeline returns nonzero photocenter shifts as a function of wavelength or dither position, those shifts define an instrument systematic floor that must be subtracted before the 35--50 microarcsecond astrophysical claims can stand.","tokens_in":12112,"feed_emoji":"🔭","tokens_out":8653,"duration_ms":82745,"temperature":0.7,"pith_summary":"FIRST-PL is a newly commissioned visible-light spectrograph on Subaru/SCExAO that feeds the corrected beam into a 19-output photonic lantern and disperses all outputs at $R\\sim3{,}000$ over 620--780 nm. The paper's central claim is that this single-telescope instrument can locate the photocenter of a stellar line to tens of microarcseconds, far below the diffraction limit, and that the measured shifts are astrophysical. On-sky results include $\\sim$50 microarcsecond precision on $\\beta$ CMi's H$\\alpha$ disk, a 0.27 mas disk asymmetry, and $\\sim$35 microarcsecond precision on Altair, whose recovered spin-axis position angle agrees with long-baseline interferometry. If the claim is right, photonic-lantern spectro-astrometry turns an 8-meter-class telescope with extreme adaptive optics into a competitor to multi-telescope interferometers for stellar rotation and disk kinematics, in integration times of tens to hundreds of seconds. The same hardware also demonstrates image reconstruction and high-contrast companion detection, and the instrument is now open to the community.","feed_headline":"Photonic lantern hits 35-microarcsecond stellar spin measurements","feed_subtitle":"Subaru's 19-port fiber spectrograph sees below the diffraction limit in seconds, rivaling interferometer arrays.","key_machinery":"The central object is the 19-port photonic lantern: an adiabatic taper from a multimode input fiber to an array of single-mode output fibers that decomposes the incoming field into orthogonal spatial modes, one per port. Because the output port fluxes depend on where the source sits in the lantern's 80 mas field, the lantern acts simultaneously as a narrow-field integral-field unit and as a co-located focal-plane wavefront sensor; dispersing the ports turns that spatial encoding into wavelength-resolved spectro-astrometry. The key calibration step is inverting this spatial response, using either passive residual tip-tilt jitter in spectro-astrometry mode or a synchronised piezo tip-tilt mirror that actively scans the beam across the lantern input in image-reconstruction modes, with simultaneous visible and near-infrared focal-plane images providing real-time telemetry of the beam position.","core_discovery":"The paper establishes that a 19-port photonic lantern fed by the Subaru/SCExAO adaptive-optics system, with each port dispersed at $R\\sim3{,}000$ over 620--780 nm, makes single-aperture spectro-astrometry precise enough to work far below $\\lambda/D$. The lantern converts the multimode input beam into 19 single-mode outputs, so a source's position and phase are encoded in the relative flux of the ports; calibrating that encoding with residual tip-tilt jitter and simultaneous focal-plane PSF telemetry turns wavelength-resolved port spectra into photocenter displacements. Commissioning results show about 50 microarcsecond precision on the H$\\alpha$ line of $\\beta$ CMi, where the recovered Keplerian velocity pattern also reveals a 0.27 mas minor-axis brightness asymmetry, and about 35 microarcsecond precision on Altair, whose H$\\alpha$ spectro-astrometry yields a spin-axis position angle of $-61.8^\\circ \\pm 2.8^\\circ$ and $v\\sin i = 220 \\pm 16$ km/s. The authors argue these measurements, taken in tens to hundreds of seconds on a single 8.2 m telescope, reach the spatial precision of long-baseline optical interferometry without its operational cost.","pith_inferences":["If the $\\sim$10 microarcsecond precision quoted for tip-tilt-modulated Altair data holds under routine observing, the practical limit of the method becomes the calibration stability of the lantern response, not photon noise; that is the number to watch as the instrument goes to open time.","A decisive stress test, not reported here, would be to scan an internal calibration point source across the lantern input with the same piezo dither pattern used on sky and demand zero recovered photocenter shift; the residual would set the true systematic floor of the technique.","The same response-matrix formalism that enables spectro-astrometry should also allow model-independent image reconstruction of sub-diffraction binaries, and the upcoming reconstruction of HIP 81126 will show whether the 70 mas binary is recovered at the expected position angle and separation."],"forward_implications":["Stellar spin-axis orientations and $v\\sin i$ values for rapidly rotating stars can be measured from a single telescope in about a minute of on-sky time, bypassing the scheduling and baseline coverage of interferometric arrays.","Circumstellar disk structure, including near-far side opacity asymmetries at the sub-milliarcsecond level, becomes accessible through wavelength-resolved photocenter shifts across emission lines.","The planned $R\\sim63{,}000$ echelle upgrade would extend the technique from broad H$\\alpha$ lines of fast rotators to narrow photospheric lines of slowly rotating stars, which is the regime needed for exoplanet-host obliquity work.","Using the lantern as a chromatic wavefront sensor could mitigate the low-wind effect on large apertures, since broad wavelength coverage avoids phase-wrapping ambiguity at the telescope spiders.","With FIRST-PL offered for open-time observations from semester 2026B, the three demonstrated modes—spectro-astrometry, on-axis image reconstruction, and off-axis high-contrast imaging—are available to external programs."],"supporting_citations":[{"why":"Establishes the single-aperture spectro-interferometry heritage on the same telescope that FIRST-PL upgrades.","marker":"[10]"},{"why":"Describes the SCExAO extreme adaptive-optics system that feeds the corrected beam to FIRST-PL.","marker":"[17]"},{"why":"Supplies the photonic-lantern principle: adiabatic multimode-to-single-mode transition with high throughput.","marker":"[18]"},{"why":"Provides the theoretical and pipeline basis for photonic-lantern spectro-astrometry and the beta CMi disk results reported here.","marker":"[22]"},{"why":"Shows that a photonic lantern acts as an all-photonic focal-plane wavefront sensor, the self-calibration mechanism FIRST-PL relies on.","marker":"[23]"},{"why":"Documents the previous visible photonic-lantern laboratory characterization and first on-sky demonstration at Subaru.","marker":"[25]"},{"why":"Presents the Altair spin-axis retrieval whose numbers this commissioning paper reproduces and summarises.","marker":"[29]"},{"why":"Provides the realistic Altair model used for comparison with the measured spin-axis parameters.","marker":"[30]"},{"why":"Provides the long-baseline interferometric imaging of Altair against which the spin-axis position angle is checked.","marker":"[31]"}],"fun_headline_variants":["Subaru's 19-port photonic lantern sees below diffraction limit","35-microarcsecond spin from a single 8.2m telescope","Lantern-fed spectrograph rivals interferometric astrometry","Subaru lantern measures Altair spin at 35 microarcsec"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The method assumes that the mapping between a star's position on the lantern input and the brightness pattern across the 19 output spectra stays identical between the calibration observation and the science observation; any drift or change in optics in between turns the tiny measured position shifts into instrument artifacts rather than real celestial structure.","fun_headline_variants_meta":{"raw":{"variants":["Subaru's 19-port photonic lantern sees below diffraction limit","35-microarcsecond spin from a single 8.2m telescope","Lantern-fed spectrograph rivals interferometric astrometry","Subaru lantern measures Altair spin at 35 microarcsec"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000786,"raw_usage":{"total_tokens":3475,"prompt_tokens":962,"completion_tokens":2513,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":578,"completion_tokens_details":{"reasoning_tokens":2439}},"tokens_in":578,"tokens_out":2513,"duration_ms":20479,"temperature":1.0,"reasoning_tokens":2439,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T18:46:22.960011+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check is to feed an unresolved calibration source through the same optics and tip-tilt dither pattern used on sky; if the pipeline returns nonzero photocenter shifts as a function of wavelength or dither position, those shifts define an instrument systematic floor that must be subtracted before the 35--50 microarcsecond astrophysical claims can stand.","supporting_citations":[{"cited_title":"Single-aperture spectro-interferometry in the visible at the subaru telescope with first: First on-sky demonstration on keho ‘oea (αlyrae) and hokulei (αaurigae),","cited_arxiv_id":null,"evidence_quote":"Establishes the single-aperture spectro-interferometry heritage on the same telescope that FIRST-PL upgrades."},{"cited_title":"The Subaru Coronagraphic Extreme Adaptive Optics System: Enabling High-Contrast Imaging on Solar-System Scales,","cited_arxiv_id":null,"evidence_quote":"Describes the SCExAO extreme adaptive-optics system that feeds the corrected beam to FIRST-PL."},{"cited_title":"The photonic lantern,","cited_arxiv_id":null,"evidence_quote":"Supplies the photonic-lantern principle: adiabatic multimode-to-single-mode transition with high throughput."},{"cited_title":"On the Potential of Spectroastrometry with Photonic Lanterns","cited_arxiv_id":"2409.09120","evidence_quote":"Provides the theoretical and pipeline basis for photonic-lantern spectro-astrometry and the beta CMi disk results reported here."},{"cited_title":"An all-photonic focal-plane wavefront sensor,","cited_arxiv_id":null,"evidence_quote":"Shows that a photonic lantern acts as an all-photonic focal-plane wavefront sensor, the self-calibration mechanism FIRST-PL relies on."},{"cited_title":"Spectroscopy using a visible photonic lantern at the Subaru Telescope: Laboratory characterization and the first on-sky demonstration on Ikiiki ( alpha Leo) and Aua (alpha Ori),","cited_arxiv_id":null,"evidence_quote":"Documents the previous visible photonic-lantern laboratory characterization and first on-sky demonstration at Subaru."},{"cited_title":"FIRST-PL: on-sky retrieval of Humu’s (Altair) spin axis using a photonic lantern on Subaru telescope,","cited_arxiv_id":null,"evidence_quote":"Presents the Altair spin-axis retrieval whose numbers this commissioning paper reproduces and summarises."},{"cited_title":"A realistic two- dimensional model of altair,","cited_arxiv_id":null,"evidence_quote":"Provides the realistic Altair model used for comparison with the measured spin-axis parameters."},{"cited_title":"Imaging the Surface of Altair,","cited_arxiv_id":null,"evidence_quote":"Provides the long-baseline interferometric imaging of Altair against which the spin-axis position angle is checked."}],"review_version":1}