pith. sign in

arxiv: 2605.23814 · v2 · pith:AGDDSEQFnew · submitted 2026-05-22 · 🌌 astro-ph.IM

Integral field spectroscopy with no IFUs: combining wide-field rotational slitless spectroscopy with tomographic reconstruction

Jerry Jun-Yan Zhang , Francesco Sinigaglia This is my paper

Pith reviewed 2026-05-25 02:39 UTC · model grok-4.3

classification 🌌 astro-ph.IM
keywords integral field spectroscopyslitless spectroscopytomographic reconstructiondatacube reconstructionspectrograph designwide-field spectroscopyrotational imaging
0
0 comments X

The pith

ROSSINI reconstructs full integral field spectroscopy datacubes from a series of rotated slitless images via tomographic inversion.

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

The paper introduces a spectrograph that captures multiple 2D slitless spectra of the same field while rotating the dispersion direction between exposures. These images are treated as projections that a post-processing tomographic algorithm inverts to recover a complete 3D datacube whose angular and spectral sampling is chosen by the user. The inversion is first written as a linear matrix problem and then solved with an iterative numerical method. Tests on simulated toy datacubes recover the input data to percent accuracy after only a few hundred iterations on ordinary computers. A reader would care because the approach promises wide-field integral field spectroscopy using simpler and cheaper hardware than conventional IFU designs.

Core claim

ROSSINI generates independent detector images of the same field by rotating the telescope or the dispersion direction of the optical element and then applies a tomographic reconstruction algorithm to produce a full IFS datacube with arbitrary angular and spectral pixelization. The problem is formulated as a linear matrix inversion that can be solved iteratively; numerical experiments show that benchmark datacubes are recovered to percent accuracy in a few hundred iterations using negligible computational resources.

What carries the argument

The iterative tomographic reconstruction algorithm that inverts the linear mapping from the unknown datacube to the set of rotated slitless detector images.

If this is right

  • The number of rotations needed is set by the relative pixelization of the datacube and the detector pixel count.
  • Reconstruction reaches percent-level accuracy on the tested benchmark datacubes.
  • Only a few hundred iterations are required for convergence.
  • The method relies on existing slitless spectrograph technologies rather than custom IFU hardware.

Where Pith is reading between the lines

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

  • Existing slitless spectrographs could be adapted for this rotational mode without major hardware changes.
  • Real observations will require separate calibration of rotation angles and optical distortions before the inversion can be applied.
  • The same rotational-projection approach might be tested on archival slitless data to check consistency with known IFU observations of the same fields.

Load-bearing premise

The tomographic algorithm recovers the true datacube from the rotated slitless images when there is no unmodeled noise, optical aberrations, or detector effects.

What would settle it

Run the reconstruction on simulated detector images that include realistic levels of noise, aberrations, and detector artifacts, then measure the residual difference from the known input datacube.

Figures

Figures reproduced from arXiv: 2605.23814 by Francesco Sinigaglia, Jerry Jun-Yan Zhang.

Figure 1
Figure 1. Figure 1: Left: On-detector images of a field with 300 stars, in the case for which the dispersion element is rotated with respect to the [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Reconstructed images of a full datacube containing 100 stars with uniform spectra, for the case in which the dispersion [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Same as Figure [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Reconstructed flux profiles of a full datacube containing 100 stars with uniform spectra, along the [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
read the original abstract

Among spectroscopic techniques, Integral Field Spectroscopy (IFS) is regarded as one of the most versatile and powerful, but it is limited by small FoVs, complex designs, and high costs. We hereby present ROSSINI: the ROtational Slitless Spectrograph for INtegral field spectroscopy and Imager, a novel spectrograph design aimed at performing IFS without IFUs but in a slitless fashion instead. The device relies on generating a series of independent detector images of the same field by rotating the whole telescope and/or the dispersion direction of the optical element, and on a postprocessing tomographic reconstruction algorithm, yielding a full IFS datacube with arbitrary angular and spectral pixelization defined by the user. We first develop the mathematical formulation of the problem and derive the solution as a linear matrix inversion. Then, we provide an interpretation of ROSSINI as a particular application of tomography and leverage this to propose a practical numerical solution based on iterative reconstruction. We test this novel conception through a series of numerical experiments: we first generate toy datacubes on the sky, then simulate the spectrograph and the corresponding detector images, and finally apply tomographic reconstruction, recovering the input datacube. The number of needed rotations can be easily computed from the relative pixelization of the datacube and the pixel number of the detector. From the numerical experiments, ROSSINI turns out to be able to reconstruct the benchmark datacubes with percent accuracy and in just a few hundred iterations, using negligible computational resources. ROSSINI is a promising way forward for future spectroscopic facilities, as it allows us to perform wide-field IFS in an efficient and cheap fashion by relying mostly on existing slitless spectrograph technologies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper claims to introduce ROSSINI, a rotational slitless spectrograph for integral field spectroscopy without IFUs. It derives a linear matrix inversion for the problem and provides a tomographic interpretation leading to an iterative reconstruction algorithm. Numerical experiments with toy datacubes show reconstruction of the input datacube with percent accuracy in a few hundred iterations using negligible computational resources, positioning it as a promising, efficient, and cheap method for wide-field IFS.

Significance. If the reconstruction accuracy holds under realistic observational conditions, this method could enable cost-effective wide-field integral field spectroscopy by building on existing slitless technologies, potentially reducing the complexity and cost of future spectroscopic facilities. The paper provides a clear mathematical formulation and tomographic view, which are strengths. No machine-checked proofs or reproducible code are mentioned, but the conceptual approach is novel in this context.

major comments (2)
  1. [Numerical experiments (abstract and main text)] The numerical experiments generate detector images using the same linear forward operator that is subsequently inverted by the matrix or iterative tomographic algorithm. This demonstrates solvability of the idealized discrete problem by construction but provides no evidence on stability when the real optical model deviates (e.g., wavelength-dependent aberrations, finite slitless PSF, or detector non-linearities). No condition numbers, singular-value spectra, or controlled model-mismatch tests are reported, which is load-bearing for the central claim of percent-level accuracy in the abstract and numerical experiments section.
  2. [Abstract and numerical experiments section] The claim that ROSSINI reconstructs benchmark datacubes 'with percent accuracy' and 'in just a few hundred iterations' lacks supporting quantitative details such as per-voxel error maps, convergence criteria, or noise models; the high-level statement alone does not substantiate the accuracy under the idealized conditions described.
minor comments (1)
  1. [Abstract] The abstract states that 'the number of needed rotations can be easily computed from the relative pixelization of the datacube and the pixel number of the detector' but does not provide or reference the explicit formula or derivation in the main text.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed report. The comments correctly identify that the current numerical experiments are idealized. We address each major point below and will revise the manuscript to improve clarity and quantitative support while preserving the paper's focus on the conceptual framework and mathematical formulation.

read point-by-point responses

  1. Referee: [Numerical experiments (abstract and main text)] The numerical experiments generate detector images using the same linear forward operator that is subsequently inverted by the matrix or iterative tomographic algorithm. This demonstrates solvability of the idealized discrete problem by construction but provides no evidence on stability when the real optical model deviates (e.g., wavelength-dependent aberrations, finite slitless PSF, or detector non-linearities). No condition numbers, singular-value spectra, or controlled model-mismatch tests are reported, which is load-bearing for the central claim of percent-level accuracy in the abstract and numerical experiments section.

    Authors: We agree that the experiments validate the method only under the assumption of an exact forward model and do not yet address robustness to realistic mismatches. This is a genuine limitation of the current work. In revision we will (i) explicitly state the idealized nature of the tests in the abstract and numerical section, (ii) add a short discussion of the system matrix condition number for the toy cases, and (iii) outline planned future work on including wavelength-dependent aberrations and finite PSF effects. Full mismatch experiments lie beyond the scope of this conceptual paper but will be noted as necessary next steps. revision: partial

  2. Referee: [Abstract and numerical experiments section] The claim that ROSSINI reconstructs benchmark datacubes 'with percent accuracy' and 'in just a few hundred iterations' lacks supporting quantitative details such as per-voxel error maps, convergence criteria, or noise models; the high-level statement alone does not substantiate the accuracy under the idealized conditions described.

    Authors: We will revise both the abstract and the numerical experiments section to include explicit quantitative metrics (mean and maximum relative error per voxel, iteration count at convergence defined by a relative residual threshold, and confirmation that the simulations are noise-free). A convergence plot and a representative per-voxel error map will be added as a new figure or panel. These changes will make the accuracy claims directly traceable to the reported experiments. revision: yes

Circularity Check

0 steps flagged

Mathematical derivation of linear inversion and tomography is independent; numerical tests are standard consistency checks

full rationale

The paper derives the linear matrix inversion directly from the forward model of rotated slitless spectroscopy and interprets the problem as tomography to motivate an iterative solver. The numerical experiments generate toy datacubes, apply the forward operator to produce detector images, and recover the input via the derived inverse; this recovers the input by the algebraic properties of the exact inverse under idealized conditions, which is a routine validation of implementation correctness rather than a self-referential definition or fitted prediction. No equations reduce to their inputs by construction in a load-bearing way, no self-citations are invoked for uniqueness or ansatzes, and the central claim rests on the first-principles formulation itself. The work is self-contained against its stated mathematical benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no explicit free parameters, axioms, or invented entities; the method is framed as a direct application of existing linear algebra and tomography techniques.

pith-pipeline@v0.9.0 · 5843 in / 1045 out tokens · 22432 ms · 2026-05-25T02:39:11.178162+00:00 · methodology

discussion (0)

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