{"id":"75b5a3c7-0033-4de6-9c84-32a5f91c975c","arxiv_id":"2509.07057","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A conceptual opto-mechanical design for SHARP, a multi-mode near-infrared spectrograph for the ELT, claims diffraction-limited performance over a 1 arcmin field with two interchangeable observing modes.","lead":"SHARP is a proposed near-infrared spectrograph for the ELT, combining a multi-object spectrograph (NEXUS) and a multi-integral-field unit (VESPER) behind the MORFEO adaptive optics system. This paper presents the conceptual optical and mechanical design, claiming it fits cryogenic and space constraints while meeting the scientific requirements of 30 mas resolution and R>1000 spectroscopy.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Encircled-energy claim rests on idealized Gaussian input PSF; realistic Airy/MORFEO PSF may drop NEXUS EE below 90% in one pixel.","rationale":"The reader's weakest assumption points to both the Gaussian PSF and the field size. I agree on the PSF but consider the field size a separate interface requirement rather than an internal modeling assumption. The load-bearing issue is the input PSF shape: it directly determines the numerical EE claim. A Gaussian input overestimates the core concentration of a real point-source image. The proposed test isolates the instrument's contribution by replacing the idealized input with a physically motivated PSF. If the EE is still ≥90%, the design is robust; if not, the headline claim is conditional on AO performance that is not yet demonstrated. This does not change the reader's CONDITIONAL verdict, so UNCHANGED is appropriate. The paper does present real structure: an all-spherical design, a compact cryostat concept, and modular mechanics, and it explicitly defers validation to future work; those are consistent with a conditional assessment.","tokens_in":12076,"tokens_out":6684,"duration_ms":75583,"concrete_test":"Recompute the EE curves of Fig. 6 (NEXUS, 2.19 µm, central and ±36\" positions) and Fig. 9 (VESPER) using a physically realistic MORFEO K-band PSF (diffraction-limited Airy core with ELT obscuration/spiders plus a seeing halo, Strehl ≈ 0.6) instead of the Gaussian input. If the NEXUS EE within one 35 mas pixel falls below 90%, the central optical-performance claim is not validated by the current simulations; if it remains ≥90%, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Sec. 3 states the EE simulations (Figs. 6 and 9) use 'a Gaussian with FWHM equal to the ELT diffraction limit' as input. A Gaussian PSF is not the diffraction-limited PSF of a telescope: the true Airy pattern (with the ELT's central obscuration and spiders) has significant flux in the first ring and wings, and the AO-corrected PSF additionally has a seeing-limited halo. At 2.19 µm, λ/D ≈ 12 mas; a 35 mas pixel subtends radius ≈1.5 λ/D, where the Airy encircled energy is only ~88% before adding NEXUS aberrations, versus >99% for the input Gaussian. Thus the claim that 'more than 90% of the flux falls within one NEXUS pixel' is primarily a property of the assumed input PSF, not a demonstrated performance of the optical design. The same input is used in Fig. 9 to conclude that VESPER's image quality is dominated by pre-SHARP optics; a realistic PSF with a halo would change that conclusion. The paper's own limitations (Sec. 5: no FEA, tolerancing, or prototyping) make this input choice load-bearing rather than cosmetic.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents the conceptual opto-mechanical design of SHARP, a near-infrared multi-mode spectrograph conceived for the ELT/MORFEO second port. SHARP consists of two subsystems: NEXUS, a multi-object spectrograph covering 0.95–2.45 µm in four simultaneous bands using 30 configurable slits and four cameras, and VESPER, a multi-integral-field unit with 12 field-selector probes, image slicers, and eight cameras. The design uses only spherical surfaces, fits in a 2 m diameter, 3 m tall cryostat, and includes mechanisms for slit configuration, field selection, and unit selection. The paper reports optical performance from ray-tracing simulations: encircled energy plots for NEXUS at 2.19 µm claim more than 90% of flux within one 35 mas pixel, and VESPER simulations are used to argue that image quality is dominated by pre-SHARP optics. The mechanical architecture is described at a conceptual level, with finite element analysis, tolerance analysis, stray-light analysis, and prototype validation deferred to future work.","tokens_in":12348,"tokens_out":3351,"duration_ms":42003,"significance":"If the quantitative performance claims are correct, the SHARP concept would demonstrate that a compact, all-spherical, multi-mode cryogenic spectrograph can simultaneously cover 0.95–2.45 µm at R>1000 over the MORFEO-corrected field, using a single 2 m-class cryostat. The modular opto-mechanical layout, the use of two moving mirrors for optical-path compensation in the field selectors, and the simultaneous four-band MOS design are genuinely useful concepts for next-generation instrumentation. However, the current performance evidence rests heavily on idealized PSF assumptions and lacks tolerance, thermal, and stray-light analyses. The potential significance is high, but the paper in its present form is a conceptual study with limited validation rather than a demonstrated design that meets all requirements.","major_comments":[{"comment":"The encircled-energy result for NEXUS is obtained with an input 'Gaussian with FWHM equal to the ELT diffraction limit.' This is not the diffraction-limited PSF of the ELT: the actual Airy pattern, including the central obscuration and spiders, has substantial flux outside the first dark ring. At 2.19 µm, λ/D is about 12 mas, and one NEXUS pixel is 35 mas; the encircled energy of a pure Airy pattern within a radius of about 1.5λ/D is approximately 88%, before any SHARP aberrations are added, whereas a Gaussian with the same FWHM contains essentially all of its flux in the same radius. Therefore the claim that 'more than 90% of the flux falls within one NEXUS pixel' is primarily a property of the assumed input PSF, not a demonstrated performance of the NEXUS optical design. Please recompute the EE with a realistic ELT/MORFEO PSF (including obscuration, spiders, and an AO residual halo) an","section":"Sec. 3, Fig. 6"},{"comment":"The conclusion that VESPER's image quality is 'dominated by the optical systems that precede SHARP' is based on comparing an ideal point source (Fig. 8) with a Gaussian input whose FWHM equals the ELT diffraction limit (Fig. 9). A real MORFEO-corrected PSF is broader and contains a seeing-limited halo, so the comparison does not establish that the pre-SHARP optics dominate in real operating conditions. The authors should repeat the comparison with a more representative MORFEO PSF, or at least quantify the effect of adding a halo component with a plausible contrast ratio, before drawing this conclusion.","section":"Sec. 3, Figs. 8–10"},{"comment":"The throughput values stated for NEXUS ('~90% in K, neglecting the grism') and VESPER ('~75% in K, neglecting the grism') are presented without any supporting calculation, transmission budget, or reference. Given that each channel includes multiple lenses, dichroics, and folding mirrors, these values are not self-evident. Provide a throughput budget with assumed coating reflectances/transmittances, or cite the source of the numbers. This is load-bearing because the scientific motivation (e.g., sensitivity to OH-line subtraction and faint continuum) depends on preserving throughput.","section":"Sec. 2.1, Sec. 2.2"},{"comment":"The paper explicitly defers finite element analysis, tolerance analysis, stray-light analysis, and prototype validation to future work. This is acceptable for a conceptual design, but the abstract's claim that the design is 'engineered to meet project specifications' is stronger than what is demonstrated. In particular, no tolerance sensitivity is presented, so it is unknown whether the all-spherical optical design remains diffraction-limited after manufacturing and cooldown. At minimum, a tolerance allocation or a sensitivity study for the most critical alignments (e.g., lens centering and despace, folding-mirror orientation) would be needed to support the 'meet specifications' wording.","section":"Sec. 5"}],"minor_comments":[{"comment":"There are several typographical inconsistencies: 'James Web Space Telescope' in the introduction (and 'James Web' in Sec. 1) should be 'James Webb Space Telescope'; decimal commas appear inconsistently in Tables 2 and 3 (e.g., '3017,1', '358,3', '20,10'), which should be unified to decimal points.","section":"Throughout"},{"comment":"The entries for camera lenses C-L5 and C-L6 list '20,10 ZNSE, SILICA' and '10,20 SILICA, ZNSE' without clarifying which thickness corresponds to which material or whether these are cemented doublets or separated elements. Please define the notation.","section":"Table 3"},{"comment":"Reference [8] is cited as a SPIE conference series but the entry contains an arXiv identifier and formatting that does not match the other references. Please convert to a consistent citation style.","section":"References"},{"comment":"The captions for the spot diagrams and encircled-energy plots do not state which field positions and wavelengths are used, nor whether the plots include the pre-SHARP optics. Adding these details would make the figures interpretable without searching the text.","section":"Figs. 6–10"},{"comment":"The sentence describing the VESPER field selector movement says the FSs 'can be deployed over a range of ∼70\" along the y-axis' and later says the probed area is 'approximately 24\"×70\"'; this geometry is not fully clear. A sketch or explicit statement of how 12 FSs (6 per module) map to the 24\"×70\" field would help.","section":"Sec. 2.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a concept design study with no prototype, no tolerance analysis, and no stray-light analysis. For a journal like JATIS, the conceptual design could be acceptable, but the optical performance claims in Sec. 3 are not robust because they assume a Gaussian PSF instead of a realistic ELT/MORFEO PSF. If the authors do not revise the simulations, I would advise against publication; the central performance claims would be unsubstantiated. The paper also relies on a companion science-case paper (Saracco et al.) for requirements, which is appropriate but should be cited more precisely. I think major revision is the right recommendation: the design is promising and the paper can be made publishable with realistic PSF inputs and a clear statement that quantitative claims are preliminary."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nYou asked what I made of the SHARP conceptual design paper. Short version: it's a real instrument proposal, with enough optical prescription and mechanical layout to be worth a serious look, but the performance numbers are more optimistic than demonstrated, and the encircled-energy plots lean on an idealized input PSF.\n\nWhat's actually new: the full opto-mechanical architecture of SHARP — 12 cameras, four dichroic bands for the MOS mode, an eight-camera mIFU with deployed pickoffs, all in a 2-m cryostat, with no aspherics. That last choice is interesting; it buys manufacturability and the spot diagrams suggest the design is well corrected across the field. The paper also lays out a sensible modular assembly plan and traces its requirements to the Saracco et al. science case. For an ELT/MORFEO second-port spectrograph, this is a useful feasibility reference.\n\nWhat it does well: it is honest about being conceptual. Section 5 explicitly defers FEA, tolerance analysis, and prototyping. The optical simulations are not fabricated; the lens prescriptions are given. The citation pattern is fine; the earlier science-case paper is not circular here.\n\nThe soft spots are real but not disqualifying at this stage. The largest is the Gaussian PSF with FWHM equal to the ELT diffraction limit used as input in Figures 6 and 9. An Airy PSF with central obscuration and spiders has a different encircled-energy curve — roughly 88% at a radius of 1.5 λ/D rather than >99% for the Gaussian. So the \"more than 90% in one pixel\" claim is partly inherited from the input, not a guaranteed instrument property. The conclusion that VESPER's image quality is dominated by pre-SHARP optics also depends on that input. I'd ask the authors to rerun the EE with a realistic MORFEO PSF or at least caveat the curves as an upper bound. The throughput numbers (90% NEXUS, 75% VESPER) appear without supporting transmission calculations; with six or more lenses plus dichroics, 90% strikes me as high. And there's no tolerance analysis, which matters for a 12-camera cryogenic system; but again, that's explicitly deferred.\n\nWho gets value: instrumentation people, especially those thinking about ELT second-port instruments and NIR multi-object/IFU spectrographs. It's not a science-results paper. As a conceptual design study, it deserves a serious referee; the verdict conditional is right. I'd send it to review, with the requirement that the PSF caveat and throughput claims be addressed before acceptance.\n\nBest,\n[You]","headline":"A genuinely useful conceptual opto-mechanical design for an ELT/MORFEO near-IR spectrograph, but the performance claims lean on an idealized Gaussian PSF and should be treated as feasibility, not proof.","tokens_in":12981,"tokens_out":3592,"would_cite":true,"duration_ms":38658,"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":"SHARP's conceptual design keeps more than 90% of a diffraction-limited source's flux inside one 35-mas pixel at 2.19 µm while covering 0.95–2.45 µm at R>1000, using an all-spherical, modular cryogenic layout.","keywords":["near-infrared spectrograph","multi-object spectroscopy","integral field unit","adaptive optics","Extremely Large Telescope","cryogenic opto-mechanics","encircled energy","configurable slit mask"],"falsifier":"Use the real MORFEO second-port PSF, from on-sky measurements or end-to-end AO simulations, and recompute the fraction of flux within a 35-mas pixel and the VESPER spot sizes; if the 90% encircled-energy figure drops below the requirement or the corrected field cannot cover 1 arcmin, the conceptual design's main performance claim is refuted.","tokens_in":11945,"feed_emoji":"🔭","tokens_out":8001,"duration_ms":83625,"temperature":0.7,"pith_summary":"This paper presents a conceptual opto-mechanical design for SHARP, a near-infrared spectrograph intended for the second port of the ELT's MORFEO adaptive-optics system. The authors claim that one cryogenic instrument can deliver both multi-object and multi-integral-field spectroscopy over 0.95–2.45 µm at spectral resolution R>1000, while resolving 30-mas structures. NEXUS splits its field into four simultaneous wavelength bands with a dedicated camera per band, so the full range is covered without reconfiguring; VESPER uses movable field-selector probes and an image slicer to map 12 small patches. Simulated encircled energy puts more than 90% of a diffraction-limited source's flux inside one 35-mas pixel at 2.19 µm, and VESPER's image quality is dominated by the incoming AO-corrected beam rather than by SHARP itself. The whole instrument fits in a 2-m-diameter cryostat with an all-spherical, modular cold structure.","feed_headline":"New spectrograph design keeps 90% of light in one 35-mas pixel","feed_subtitle":"SHARP covers 0.95–2.45 µm at R>1000 with all-spherical optics inside a compact 2-meter cryostat.","key_machinery":"The central object is the optical train from the MORFEO second port through a Unit Selector System that switches between two front ends. NEXUS uses a configurable slit system and three dichroics to split light into four bands, each feeding a dedicated camera with a grism wheel; VESPER uses 12 movable field-selector probes, arranged in two modules, whose constant optical path is maintained by pairs of mirrors moving at half the probe displacement, feeding an image slicer with 72 micro-mirrors per stripe and four cameras per module. These mechanisms, plus an all-spherical lens prescription and modular cast-aluminum cold structure, carry the argument that the instrument can meet resolution and","core_discovery":"The authors claim that a single cryogenic instrument, SHARP, can exploit the AO-corrected field of MORFEO at the ELT to perform both multi-object and multi-integral-field spectroscopy in the near-infrared. The design splits NEXUS light into four simultaneous bands, each with its own camera and grism wheel, so the full 0.95–2.45 µm range is covered without reconfiguring; VESPER uses 12 movable field-selector probes and an image slicer to feed eight cameras. Encircled-energy simulations at 2.19 µm show >90% of the flux from a diffraction-limited Gaussian source falls inside one 35-mas NEXUS pixel, and VESPER spot diagrams are dominated by the input AO optics rather than SHARP itself, so both m","pith_inferences":["The >90%-in-one-pixel figure is an end-to-end estimate for a Gaussian input PSF; a real AO PSF with broader wings will reduce that fraction even if the FWHM matches, so the margin should be tested against realistic PSF profiles.","The design implies a large detector count (16 2k×2k for NEXUS and 8 4k×4k for VESPER); the paper does not estimate detector cost, data volume, or readout time, which could dominate project feasibility.","Removing the ADC for VESPER observations is proposed to gain sensitivity, but the gain is not quantified; computing it would be a natural next step.","If the modular all-spherical design meets its goals at the ELT port, the same architecture could be offered to other ELT-class MCAO instruments with similar focal ratios, though the 1×1 arcmin corrected-field assumption would need rechecking."],"forward_implications":["A single NEXUS exposure can cover the full 0.95–2.45 µm range without swapping gratings, avoiding the configuration changes needed by existing near-infrared multi-object spectrographs.","Up to 30 configurable slits can be placed on a 1.2×1.2 arcmin field, giving multiplexed spectroscopy at R≈300, 2000, or 6000 depending on grism choice.","VESPER's 12 probes return spatially resolved spectra over a 24×70 arcsec patrol area with 0.031-arcsec slices, matching the scales of giant molecular clouds at high redshift.","The 2-m cryostat and modular cast-aluminum structure keep cryogenic power and assembly effort manageable, with front/rear cap access for maintenance.","Because VESPER's images are limited by the input AO PSF, further improvement of SHARP's own optics will not sharpen VESPER data until the upstream adaptive-optics system improves."],"supporting_citations":[{"why":"Defines MORFEO as the multi-conjugate adaptive-optics system whose corrected field and second port feed SHARP; supplies the input assumptions for field size and PSF.","marker":"1,2"},{"why":"MICADO is the first-light imager sharing the NGS unit design that SHARP adopts without the SCAO module, anchoring the wavefront-sensing concept.","marker":"3"},{"why":"NIRSpec is the JWST spectrograph whose wavelength and resolution coverage SHARP is designed to surpass, setting the comparison baseline.","marker":"6"},{"why":"Sets the science requirements (2.45 µm limit, 30 mas resolution, R>1000, 1×1 arcmin field) that drive the SHARP design.","marker":"8"},{"why":"EMIR at the GTC is an existing near-infrared multi-object spectrograph that must switch configurations between wavelength bands, justifying NEXUS's simultaneous four-band design.","marker":"9-11"},{"why":"The mechanical slit-mask mechanism provides the configurable-slit concept adopted for NEXUS's Configurable Slit System.","marker":"13"},{"why":"MOSFIRE is the comparable near-infrared multi-object spectrograph whose configurable slit unit and grating design are the direct heritage and comparison for NEXUS.","marker":"14"},{"why":"The configurable slit-mask unit for MOSFIRE was integrated and tested, validating the CSS mechanism approach for cryogenic multi-object spectroscopy.","marker":"15"}],"fun_headline_variants":["SHARP spectrograph packs 90% light into one 35-mas pixel","One cryostat, two spectrographs: SHARP for ELT","Next-gen spectrograph SHARP keeps 90% light in a single pixel","SHARP design: 90% encircled energy in one pixel for ELT","Compact cryostat SHARP delivers multi-object and IFU spectroscopy"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The performance numbers assume the adaptive-optics system actually delivers a star image as sharp as the ELT diffraction limit across at least a 1×1 arcmin field; if the real corrected image is softer or the corrected field smaller, the claimed resolution and encircled energy would not be reached.","fun_headline_variants_meta":{"raw":{"variants":["SHARP spectrograph packs 90% light into one 35-mas pixel","One cryostat, two spectrographs: SHARP for ELT","Next-gen spectrograph SHARP keeps 90% light in a single pixel","SHARP design: 90% encircled energy in one pixel for ELT","Compact cryostat SHARP delivers multi-object and IFU spectroscopy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000193,"raw_usage":{"total_tokens":1190,"prompt_tokens":748,"completion_tokens":442,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":492,"completion_tokens_details":{"reasoning_tokens":340}},"tokens_in":492,"tokens_out":442,"duration_ms":4541,"temperature":1.0,"reasoning_tokens":340,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T22:50:01.166174+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Use the real MORFEO second-port PSF, from on-sky measurements or end-to-end AO simulations, and recompute the fraction of flux within a 35-mas pixel and the VESPER spot sizes; if the 90% encircled-energy figure drops below the requirement or the corrected field cannot cover 1 arcmin, the conceptual design's main performance claim is refuted.","supporting_citations":[{"cited_title":"MICADO: The Multi-Adaptive Optics Camera for Deep Observations,","cited_arxiv_id":null,"evidence_quote":"MICADO is the first-light imager sharing the NGS unit design that SHARP adopts without the SCAO module, anchoring the wavefront-sensing concept."},{"cited_title":"The Near-Infrared Spectrograph (NIRSpec) on the JamesWebb Space Telescope. I. Overview of the instrument and its capabilities,","cited_arxiv_id":null,"evidence_quote":"NIRSpec is the JWST spectrograph whose wavelength and resolution coverage SHARP is designed to surpass, setting the comparison baseline."},{"cited_title":"Mechanical slit mask mechanism for the James Webb Space Telescope spectrometer,","cited_arxiv_id":null,"evidence_quote":"The mechanical slit-mask mechanism provides the configurable-slit concept adopted for NEXUS's Configurable Slit System."},{"cited_title":"Design and development of MOSFIRE, the Multi-Object Spectrometer For Infra-Red Exploration at the Keck Observatory","cited_arxiv_id":null,"evidence_quote":"MOSFIRE is the comparable near-infrared multi-object spectrograph whose configurable slit unit and grating design are the direct heritage and comparison for NEXUS."},{"cited_title":"Configurable slit-mask unit of the Multi- Object Spectrometer for Infra-Red Exploration for the Keck telescope: Integration and Tests","cited_arxiv_id":null,"evidence_quote":"The configurable slit-mask unit for MOSFIRE was integrated and tested, validating the CSS mechanism approach for cryogenic multi-object spectroscopy."}],"review_version":1}