Marcelo Tala Pinto, Marshall C. Johnson, Julia Brady, Xavier Lesley-Saldaña, Jonathan Crass, Allie Renshaw, Michael Engelman, Brian Sands
Competent, honestly-reported instrument characterization milestone with a transferable focusing methodology, but narrow in scope and largely tied to one specific instrument.
iLocater is a diffraction-limited, fiber-fed spectrograph designed for the Large Binocular Telescope (LBT), which targets high-precision radial velocity measurements in the near-infrared. Prior to deployment, comprehensive laboratory characterization was essential to validate instrument performance and inform alignment strategies. This paper presents results from four key areas of lab characterization: (1) adjustment and optimization of detector orientation to optimize spectrum alignment with the detector pixel grid across the focal plane; (2) the design and installation of a LED illumination source to enable high-fidelity flat-fields; (3) a model-based focusing methodology using OpticStudio image simulations to optimize the optical alignment of the spectrograph; and (4) assessment of instrument mechanical and optical stability under laboratory conditions. Together, these efforts established baseline performance metrics and demonstrated instrument readiness for delivery and on-sky commissioning.
Paper type: Instrumentation/engineering characterization paper (SPIE conference proceedings).
This paper reports the pre-deployment laboratory characterization of iLocater, a diffraction-limited, single-mode-fiber-fed near-infrared echelle spectrograph (R∼205,000) for the Large Binocular Telescope aimed at extreme-precision radial velocity (EPRV) detection of low-mass exoplanets. The paper documents four discrete engineering activities: (1) detector clocking correction (~2° rotation to align spectral orders with the pixel grid); (2) design and installation of an LED-based flat-field illumination system; (3) a model-based focusing methodology using Zemax OpticStudio simulations (via `zospy`) that incorporates thermal-expansion effects of Invar and silicon components; and (4) a two-week thermal/optical stability assessment demonstrating sub-1 m/s RV drift (0.86 m/s rms for the central fiber) correlated with ~1.5 mK temperature drift. The primary novelty is the model-based, simulation-guided focusing approach that reduces the number of costly cryogenic thermal cycles required for alignment—a genuine practical contribution for cryogenic instrument commissioning.
The methods are sound and appropriate for the engineering task. The focusing approach builds a synthetic detector image by combining OpticStudio-computed PSFs positioned at model-derived centroids, then fits quadratic sampling-vs-focus parabolas across a grid of temperatures and detector tilts, and matches measured UNe-lamp sampling from four cooldowns to converge to best focus. This is a legitimate, well-conceived procedure. The stability analysis uses standard, well-established techniques (optimal extraction, Gaussian line-centroiding, inverse-variance weighting, per-line velocity conversion). The paper honestly flags limitations: the preliminary wavelength solution is accurate only to ~0.02 nm (10–20 pixels) and inadequate for direct cross-correlation, and the top/bottom fiber flux-splitting artifact is acknowledged and attributed to the lab configuration. However, as a conference proceedings, the rigor is at the level of demonstration rather than exhaustive validation—there are no formal ablations, limited statistical treatment, and the stability run spans only two weeks under lab (not on-sky) conditions.
Impact is concentrated within the EPRV instrumentation community. The demonstrated sub-1 m/s laboratory stability and the successful diffraction-limited, single-mode approach are relevant to a small but strategically important subfield pursuing Earth-analog detection around M dwarfs. The model-based focusing methodology is transferable in principle to other cryogenic spectrographs (HPF, NEID, SPIRou, CRIRES+ heritage), and could inform alignment strategies for future single-mode instruments and ELT-class spectrographs. That said, this specific paper is a milestone report for one instrument rather than a methodological framework paper; its findings are largely tied to iLocater's particular optical/mechanical configuration. Real-world "application" here is the instrument itself, which is genuinely valuable, but the paper documents a step rather than delivering a reusable primitive.
EPRV in the NIR for M-dwarf planet detection is an active, high-priority area (endorsed by decadal science goals for finding habitable-zone planets). Diffraction-limited, single-mode-fiber spectroscopy is an emerging paradigm that iLocater helped pioneer, so the work is timely. The paper is part of a coordinated set of 2026 SPIE papers (calibration system, commissioning) marking iLocater's transition to on-sky operation.
Strengths: Clear practical value of the simulation-guided focusing (reduces expensive cryogenic cycles); honest reporting of limitations; concrete, quantitative stability results (0.86 m/s rms at 1.5 mK); demonstrated R∼205,000 matching design; use of open tools (`zospy`, `HxRGproc`). The multi-institution collaboration and integration with adaptive optics reflect a mature, well-supported program.
Limitations: Narrow scope—this is an instrument-status/characterization paper with limited generalizable insight. No single result overturns or challenges prior belief. The four activities are somewhat disconnected engineering vignettes rather than a unified scientific advance. Reproducibility outside the iLocater team is low: results depend on the specific hardware, cooldown history, and proprietary/instrument-specific configurations; no code or data release is mentioned (though underlying tools are public). The stability demonstration, while promising, is preliminary (limited wavelength solution, lab-only, 2 weeks) and does not yet demonstrate on-sky EPRV performance. The reported "improvement in PSF quality" from focusing is stated qualitatively rather than quantified in the conclusions.
This is a typical, competently executed SPIE instrumentation proceedings paper—valuable to the project and its immediate community, likely cited by subsequent iLocater papers and by teams building comparable single-mode NIR spectrographs, but unlikely to achieve broad or cross-disciplinary influence. The `zospy`-based synthetic-image focusing workflow is the element most likely to be adopted elsewhere. Resource intensity is high (LBT-class instrument, cryostat, multi-institution, NSF/NASA funding), which limits who can build directly on it, though the modeling methodology itself is accessible.
Generated Jul 31, 2026
Competent, honestly-reported instrument characterization milestone with a transferable focusing methodology, but narrow in scope and largely tied to one specific instrument.