Pith. sign in

REVIEW 3 major objections 6 minor 10 references

SPRINT tracks and corrects DM/WFS mis-registrations on-sky at LBT, restoring the AO system to its nominal alignment.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

A simplified SPRINT loop at LBT recovers and corrects DM/WFS rotation and shift mis-registrations in daytime and on-sky tests, including an unplanned event that raised Strehl from 27% to 67%.

T0 review reviewed 2026-07-31 challenge →

load-bearing objection First on-sky closed-loop SPRINT at LBT works for rotation; shift still has an unresolved scale error the authors already flag. the 3 major comments →

arxiv 2607.27941 v1 pith:GK3QUKTX submitted 2026-07-30 astro-ph.IM

DM/WFS mis-registration tracking: Implementation and on-sky validation of SPRINT at LBT

Ben Buky , C\'edric Ta\"issir H\'eritier , Fabio Rossi , Juan Carlos Guerra , Charlotte Z. Bond , Noah Schwartz , Guido Agapito , Enrico Pinna
show 2 more authors
Sam Ragland Jean-Fran\c{c}ois Sauvage
This is my paper
classification astro-ph.IM
keywords Mis-registrationPyramid wavefront sensorOn-sky testingLBTELTAdaptive opticsSPRINTDeformable mirror
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Telescopes that put the deformable mirror inside the telescope itself leave a long optical path between that mirror and the wavefront sensor. Small drifts in rotation or lateral shift between the two destroy the adaptive-optics loop and block diffraction-limited images. This paper shows that a simplified, single-iteration version of the SPRINT estimator, driven by a tiny probe already present for optical-gain tracking, can measure those drifts in real time and send corrections to the de-rotator and camera lens. Daytime and on-sky tests at the Large Binocular Telescope recover injected mis-registrations and return the system to its calibrated state; an unplanned large rotation event raised the near-infrared Strehl ratio from 27 % to 67 %. The result supplies the first on-sky proof that SPRINT can keep an integrated-DM adaptive-optics system registered during science observations, the same problem that will face every ELT instrument.

Core claim

A reduced-complexity, single-iteration SPRINT loop running as an auxiliary process on the LBT real-time computer accurately identifies and corrects both rotation and shift mis-registrations between the adaptive secondary and the pyramid wavefront sensor, returning the system to its nominal registration in daytime and on-sky tests and restoring science-image quality after an unplanned large rotation offset.

What carries the argument

SPRINT: a pseudo-synthetic model of the AO system that infers rotation and shift parameters from the demodulated wavefront-sensor response to a small, already-present sinusoidal probe (KL mode 30 at 80 Hz, ~10 nm).

Load-bearing premise

A single-pass estimator that ignores optical-gain and mis-registration-induced gain changes still produces corrections accurate enough when driven by an unoptimized existing probe.

What would settle it

Repeat the on-sky shift-injection test under the same conditions: if the factor-of-two overestimate persists or if closed-loop Strehl fails to recover after deliberate mis-registration, the single-iteration estimator is inadequate.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • LBT can now run a continuous SPRINT auxiliary loop to keep SOUL registered in both shift and rotation during science.
  • The same single-iteration code path is ready for integration into ELT instruments that already selected SPRINT as baseline.
  • Bootstrapping an AO loop from tip/tilt-only while SPRINT removes large mis-registrations becomes a practical commissioning strategy.
  • Existing optical-gain probe signals can double as mis-registration trackers, avoiding extra DM disturbance.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The on-sky factor-of-two shift bias may be a fixed scale error in the sensitivity matrices; a one-time empirical recalibration could remove it without restoring full iterative gain estimation.
  • If the method remains stable at 300–500 frames per estimate, sub-second update rates become feasible for faster ELT flexure.
  • Success with an unoptimized KL-30 probe suggests many current AO systems already possess usable probe signals for SPRINT-like tracking.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The manuscript reports the implementation of a reduced-complexity, single-iteration version of SPRINT as an auxiliary loop in the LBT real-time computer, and its daytime and on-sky validation for tracking rotation and shift DM/WFS mis-registrations on the SOUL pyramid-WFS system. A pre-existing KL-mode-30 / 80 Hz / ~10 nm probe (already used for optical-gain tracking) is demodulated from WFS slopes; offline sensitivity matrices from a pseudo-synthetic model yield mis-registration estimates that are applied to the de-rotator and camera lens. Daytime tests recover commanded rotation and y-shift (including with the AO loop closed only on tip/tilt). On-sky, rotation is recovered across 300/500/1000-frame batches (>95% of estimates within the 0.5° target), and an unplanned rotation event restores LUCI Strehl from 27% to 67%. An on-sky x-shift test recovers the correct sign/direction but overestimates the commanded offset by a factor of ~2.

Significance. This is the first on-sky validation of SPRINT and a concrete operational path for telescopes with integrated DMs (LBT ASM; ELT M4). The unplanned-event Strehl recovery is strong external evidence that the estimator can detect and correct a real mis-registration under science conditions. Reusing an existing optical-gain probe and a Python-2.7 single-iteration implementation (no OOPAO dependency) are practical strengths for observatory deployment. If the remaining shift-scale issue is resolved or bounded, the work materially de-risks SPRINT as the baseline mis-registration tracker for ELT instruments such as HARMONI.

major comments (3)
  1. [Sec. 4, Fig. 6; Sec. 5] Sec. 4 and Fig. 6: the on-sky x-shift test shows a persistent factor-of-two overestimate of the commanded camera-lens offset that is absent in the corresponding daytime results (Sec. 3). The abstract and Sec. 5 claim that SPRINT “accurately identify and correct for shift and rotation” and returns the system to its nominal configuration “in all cases,” and quote a target of 10% of a subaperture. With loop gain 0.6, a ×2 scale error changes closed-loop overshoot/settling and means the published shift recovery is not yet shown to meet that target. The discrepancy must be diagnosed (sensitivity-matrix scaling, optical-gain coupling into the demodulated KL-30 amplitude, pupil geometry, or commanded vs. true lens motion) or the accuracy claim for shift must be explicitly qualified pending further tests.
  2. [Sec. 2] Sec. 2: the LBT implementation is deliberately single-iteration and “assumes no gain variations are present,” dropping the iterative gain update of Héritier et al. (2021) and reusing an unoptimized KL-30 probe. For a pyramid WFS, both optical gain and mis-registration-induced gain scale the demodulated mode-30 response that feeds the estimator; a wrong scale in the offline sensitivity matrices produces exactly a multiplicative bias in the shift estimate. The paper should quantify (even with a simple simulation or daytime gain-sweep) how large a gain mismatch can be before the shift estimate exceeds the 10%-subaperture target, and state whether the observed on-sky ×2 bias is consistent with that analysis. Without this, the joint shift+rotation accuracy claim and the ELT-readiness statement rest on an unresolved model assumption.
  3. [Sec. 4; Sec. 5] Sec. 4–5: the tip/tilt-only bootstrap demonstration (important for ELT closed-loop start-up from a severely mis-registered state) is shown only in daytime (Fig. 4); the on-sky attempt is reported as inconclusive under 1.7" seeing. The conclusions correctly flag this as future work, but the present wording that the daytime result is “a significant result in the context of using SPRINT for bootstrapping purposes at the ELT” should be tempered until on-sky confirmation exists, or the limitation should be stated more prominently in the abstract/conclusions.
minor comments (6)
  1. [Fig. 3] Fig. 3 central panel: the modal PSD spike at mode 30 is identified in the text as the injected probe, but the figure caption does not; a brief call-out would help readers.
  2. [Fig. 5] Fig. 5: clarify in the caption how sky-rotation offsets were subtracted from the “Offset on De-Rotator” traces so that nominal remains 0°, and note which panel used the standard LBT sky-rotation tracker (saw-tooth) versus SPRINT control before injection.
  3. [Table 1; Sec. 2] Table 1 and Sec. 2: state the physical meaning of the 10%/0.5° target accuracies in actuator or subaperture units for the LBT ASM/SOUL geometry so readers can compare to ELT-scale requirements.
  4. [Throughout] Typographical consistency: “IMPLEMENT A TION”, “DA YTIME”, “RESUL TS” and similar spaced capitals in section headings should be normalized; “Karhunen-Lo` eve” encoding; “python 2.7” → “Python 2.7”.
  5. [Sec. 1] Sec. 1: magnification/anamorphosis are stated as trackable by SPRINT but “have not been implemented here”; a one-sentence note on whether the single-iteration LBT code path could support them without OOPAO would be useful for ELT readers.
  6. [References] References: ensure Héritier et al. 2021 is cited consistently as the algorithmic source when contrasting iterative vs. single-iteration performance.

Circularity Check

0 steps flagged

No significant circularity: experimental validation against independently commanded offsets and external Strehl, not a self-derived prediction chain.

full rationale

This is an implementation-and-on-sky-validation paper, not a first-principles derivation. Mis-registration ground truth is set by commanded de-rotator and camera-lens offsets (and, in one case, by an independent LUCI Strehl measurement), which are external to the SPRINT estimator. The algorithm is taken from Héritier et al. 2021 and deliberately simplified to a single-iteration form that assumes no gain variations; that is an engineering choice with acknowledged accuracy cost, not a claim that the estimates are derived from pure theory. Sensitivity matrices are generated offline from a pseudo-synthetic model tuned to the current interaction matrix, then used to invert demodulated KL-30 responses—standard model-based estimation, not fitting a parameter and relabeling it as a prediction. Overlapping authorship on the method citation is normal for a follow-on implementation paper and is not load-bearing for the empirical recovery claims. No step reduces a claimed prediction to its own inputs by construction. Score 0; steps empty.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 0 invented entities

The central claim rests on standard AO control assumptions, the prior SPRINT algorithm, and several engineering choices (probe mode, single iteration, loop gain, frame count) that are free parameters of the implementation rather than derived quantities. No new physical entities are postulated.

free parameters (4)
  • SPRINT loop gain = 0.6 (default)
    Hand-chosen feedback gain on applied mis-registration corrections; not derived from a stability proof in this paper.
  • Frames per SPRINT estimate (n) = 300 / 500 / 1000
    Integration length chosen empirically; accuracy visibly degrades at 300 frames on-sky.
  • Perturbation mode, amplitude, frequency = KL mode 30, 10–15 nm, 80 Hz
    Inherited from existing optical-gain tracking; not optimized for mis-registration sensitivity in this work.
  • Target accuracy thresholds = 10% of a subaperture; 0.5 degrees
    Stated acceptance criteria used to judge success; engineering targets, not measured optima.
axioms (5)
  • ad hoc to paper A single iteration of SPRINT with fixed calibrated interaction matrix and no explicit gain-variation estimation is sufficient for operational mis-registration tracking.
    Sec. 2 explicitly drops the iterative gain update of Héritier et al. 2021 to shrink the LBT codebase; accuracy reduction is acknowledged but not bounded analytically.
  • domain assumption Small sinusoidal DM probes (~10 nm) produce a near-imperceptible science impact while yielding usable demodulated WFS signals.
    Standard invasive-calibration assumption in AO; used throughout Secs. 2–4.
  • domain assumption Commanded de-rotator and camera-lens offsets are accurate ground truth for injected rotation and shift.
    All recovery plots compare SPRINT estimates to these mechanism readbacks (Figs. 3–6).
  • domain assumption Pseudo-synthetic sensitivity matrices generated offline from the current interaction matrix adequately linearize rotation and shift responses near the operating point.
    Core SPRINT modeling step (Sec. 2); inherited from prior SPRINT literature.
  • ad hoc to paper Linear interpolation of skipped AO telemetry frames plus a local frequency scan recovers a usable demodulated KL-30 signal.
    Sec. 2 mitigation specific to LBT telemetry; not formally characterized for bias.

reviewed 2026-07-31 · how reviews work

0 comments
Cite this review

Pith. "Pith review of DM/WFS mis-registration tracking: Implementation and on-sky validation of SPRINT at LBT." pith.science (2026). https://pith.science/paper/GK3QUKTX

@misc{pith2026260727941,
  author       = {Pith},
  title        = {Pith review of: DM/WFS mis-registration tracking: Implementation and on-sky validation of SPRINT at LBT},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GK3QUKTX}},
  note         = {Machine review of arXiv:2607.27941}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The advent of telescopes with an integrated deformable mirror (DM) presents new challenges for adaptive optics (AO) systems. The alignment between the DM and wavefront sensor (WFS) is expected to regularly evolve during operations due to their large separation. Without tracking and correction, these mis-registrations between the DM and WFS lead to loop instability, preventing diffraction limited performance from being realised. SPRINT\cite{heritier2021} provides an approach to track these mis-registrations during observations. Rotation, shift, and magnification mis-registrations can all be recovered. The Large Binocular Telescope (LBT) currently lacks an operational solution for tracking these mis-registrations, while SPRINT has been selected as the baseline approach for several instruments on the forthcoming Extremely Large Telescope (ELT). We report on the implementation of SPRINT into the LBT real time computer and present experimental results from both daytime and on-sky testing to validate the method.

Figures

Figures reproduced from arXiv: 2607.27941 by Ben Buky, C\'edric Ta\"issir H\'eritier, Charlotte Z. Bond, Enrico Pinna, Fabio Rossi, Guido Agapito, Jean-Fran\c{c}ois Sauvage, Juan Carlos Guerra, Noah Schwartz, Sam Ragland.

Figure 1
Figure 1. Figure 1: A simplified optical layout of the SOUL instrument at LBT. The pupil de-rotator and camera lens allow for [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The key components of the SPRINT implementation at LBT. The model is set up offline, whilst the perturbation [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Results from injecting rotation (left) and y-shift (right) mis-registrations in daytime. The estimates obtained [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: SPRINT rotation tracking results in daytime with the AO loop closed only on tip and tilt. In the left plot [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: On-sky SPRINT results from the injection of rotation mis-registrations. The number of frames per SPRINT [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: On-sky SPRINT results from the injection of a shift mis-registration. The SPRINT estimates are plotted [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: On-sky SPRINT results from an unplanned rotation mis-registration event. The left plot is the mis-registration [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

10 extracted references

  1. [1]

    Sprint, system parameters recurrent invasive tracking: a fast and least-cost online calibration strategy for adaptive optics,

    H´ eritier, C. T., Fusco, T., Oberti, S., Neichel, B., Esposito, S., and Madec, P.-Y., “Sprint, system parameters recurrent invasive tracking: a fast and least-cost online calibration strategy for adaptive optics,”Monthly Notices of the Royal Astronomical Society504, 4274–4290 (May 2021)

    2021

  2. [2]

    Specifications and design of the E-ELT M4 adaptive unit,

    Vernet, E., Cayrel, M., Hubin, N., Mueller, M., Biasi, R., Gallieni, D., and Tintori, M., “Specifications and design of the E-ELT M4 adaptive unit,” in [Adaptive Optics Systems III],Proc. SPIE8447, 844761 (2012)

    2012

  3. [3]

    The adaptive secondary mirror for the Large Binocular Telescope: optical acceptance test and preliminary on-sky commissioning results,

    Riccardi, A., Xompero, M., Briguglio, R., Quir´ os-Pacheco, F., Busoni, L., Fini, L., Puglisi, A., Esposito, S., Arcidiacono, C., Pinna, E., Ranfagni, P., Salinari, P., Brusa, G., Demers, R., Biasi, R., and Gallieni, D., “The adaptive secondary mirror for the Large Binocular Telescope: optical acceptance test and preliminary on-sky commissioning results,”...

    2010

  4. [4]

    ESO adaptive optics facility,

    Arsenault, R., Madec, P.-Y., Hubin, N., Paufique, J., Stroebele, S., Soenke, C., Donaldson, R., Fedrigo, E., Oberti, S., Tordo, S., Downing, M., Kiekebusch, M., Conzelmann, R., Duchateau, M., Jost, A., Hackenberg, W., Bonaccini Calia, D., Delabre, B., Stuik, R., Biasi, R., Gallieni, D., Lazzarini, P., Lelouarn, M., and Glindeman, A., “ESO adaptive optics ...

    2008

  5. [5]

    Estimation of the lateral mis-registrations of the gravity+ adaptive optics system,

    Berdeu, A., Bonnet, H., Le Bouquin, J.-B., ´Edouard, C., Gomes, T., Shchekaturov, P., Dembet, R., Pau- mard, T., Oberti, S., Kolb, J., Millour, F., Berio, P., Lai, O., Eisenhauer, F., Garcia, P., Straubmeier, C., Kreidberg, L., H¨ onig, S. F., and Defr` ere, D., “Estimation of the lateral mis-registrations of the gravity+ adaptive optics system,”Astronomy...

    2024

  6. [6]

    HARMONI at ELT: wavefront control in SCAO mode,

    Bond, C. Z., Sauvage, J.-F., Schwartz, N., Levraud, N., Chambouleyron, V., Correia, C., Fusco, T., and Neichel, B., “HARMONI at ELT: wavefront control in SCAO mode,” in [Adaptive Optics Systems VIII], Proc. SPIE12185(2022)

    2022

  7. [7]

    SOUL: the Single conjugated adaptive Optics Upgrade for LBT,

    Pinna, E., Esposito, S., Hinz, P., Agapito, G., Bonaglia, M., Puglisi, A., Xompero, M., Riccardi, A., Briguglio, R., Arcidiacono, C., Carbonaro, L., Fini, L., Montoya, M., and Durney, O., “SOUL: the Single conjugated adaptive Optics Upgrade for LBT,” in [Adaptive Optics Systems V],Proc. SPIE9909, 99093V (2016)

    2016

  8. [8]

    Oopao: Object oriented python adaptive optics,

    H´ eritier, C. T., V´ erinaud, C., and Correia, C. M., “Oopao: Object oriented python adaptive optics,” AO4ELT7(2023)

    2023

  9. [9]

    Soul at lbt: commissioning results, science and future.,

    Pinna, E., Rossi, F., Agapito, G., Puglisi, A., Plantet, C., Ghose, E., Matthieu, B., Bonaglia, M., Briguglio, R., Brusa, G., Carbonaro, L., Cavallaro, A., Christou, J., Durney, O., Ertel, S., Esposito, S., Grani, P., Hinz, P., Guerra, J. C., Lefebvre, M., Mazzoni, T., Mechtley, B., Miller, D. L., Montoya, M., Power, J., Rothberg, B., Taylor, G., Vaz, A.,...

    2023

  10. [10]

    On- sky correction of non-common path aberration with the pyramid wavefront sensor,

    Esposito, S., Puglisi, A., Pinna, E., Agapito, G., Quir´ os-Pacheco, F., V´ eran, J.-P., and Herriot, G., “On- sky correction of non-common path aberration with the pyramid wavefront sensor,”Astronomy & Astro- physics636, A88 (2020)

    2020

This paper was first reviewed by grok-4.5 on July 31, 2026.