arxiv: 2604.24875 · v1 · submitted 2026-04-27 · 🌌 astro-ph.IM · astro-ph.GA

Recognition: unknown

Joint Multiband Photometry with crowdsource

Jayashree Behera , Edward F. Schlafly , Aaron M. Meisner , Lucas Napolitano

Authors on Pith no claims yet

Pith reviewed 2026-05-07 17:11 UTC · model grok-4.3

classification 🌌 astro-ph.IM astro-ph.GA

keywords multiband photometrycrowdsourcecrowded fieldsWISE imagingjoint fittingflux consistencypositional scattersource detection

0 comments

The pith

Joint multiband fitting in crowdsource shares source positions across bands to improve flux consistency and cut positional scatter.

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

The paper extends the crowdsource pipeline so that sources are fitted simultaneously in multiple bands rather than independently. It requires the same sources to occupy identical locations in every band while letting only their fluxes change. This joint approach also pools all images to detect faint sources with adjustable weighting by bandpass. A reader would care because separate per-band fits often produce mismatched positions and fluxes that degrade the reliability of crowded-field catalogs. The authors test the method on WISE W1 and W2 images and report measurable gains in agreement and stability.

Core claim

The multiband extension performs simultaneous fitting across bands by constraining sources to identical sky positions while allowing fluxes to vary freely; all available images contribute to source detection with configurable weighting. When applied to unWISE coadded tiles in both sparse and crowded regions, the joint fit produces more consistent fluxes and smaller band-to-band positional differences than independent single-band processing. The framework supplies a general foundation for constructing multiband crowded-field catalogs.

What carries the argument

The mathematical formulation of the multiband fit that enforces shared source locations across bands while permitting independent fluxes.

If this is right

Photometric catalogs from crowded fields gain higher internal consistency between bands.
Band-to-band positional scatter decreases relative to separate fits.
Faint-source detection improves by combining information from all bands with spectral weighting.
The same shared-position constraint supplies a reusable basis for other multiband crowded-field pipelines.

Where Pith is reading between the lines

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

The fixed-position assumption could support more stable color measurements for source classification across bands.
Surveys with many simultaneous bands would likely see larger reductions in catalog scatter in dense fields.
The joint framework could be extended to time-series data to separate positional stability from flux variability.

Load-bearing premise

Sources occupy exactly the same positions in every band, differing only in their measured fluxes.

What would settle it

Running the joint versus independent fits on the same set of unWISE W1-W2 tiles and finding no reduction in either flux scatter or band-to-band positional differences would falsify the performance improvement.

Figures

Figures reproduced from arXiv: 2604.24875 by Aaron M. Meisner, Edward F. Schlafly, Jayashree Behera, Lucas Napolitano.

**Figure 1.** Figure 1: Illustration of single-band versus multiband source catalog in a representative high-latitude unWISE field (1497p015). Left: W1 image cutout. Middle: Source positions recovered from independent single-band fits in W1 (blue) and W2 (red), showing band-dependent centroid scatter and missed faint detections in W2. Right: Source positions recovered from the multiband fit, yielding a single, self-consistent set… view at source ↗

**Figure 2.** Figure 2: Difference between multiband and single-band WISE magnitudes as a function of single-band magnitude for W1 (left) and W2 (right). Bright sources show excellent agreement between the two methods, while systematic differences emerge at the faint end, most prominently in W2. The negative shift in faint W2 magnitudes indicates that the sky level was underestimated in the single-band W2 photometry; including th… view at source ↗

**Figure 4.** Figure 4: Difference in source counts between the multiband and single-band catalogs in (W1 − W2) color versus W1 magnitude for a representative Galactic bulge unWISE field (2602m273). The multiband fit shows a strong excess along the main stellar locus and at the faint end, indicating improved source recovery in crowded conditions. using a 6′′ matching radius. Of these ∼3,850 new multiband detections, approximately… view at source ↗

**Figure 5.** Figure 5: Two-dimensional completeness in the COSMOS field using Spitzer COSMOS as a deeper reference catalog. Left: completeness for single-band WISE photometry. Middle: completeness for multiband WISE photometry. Right: difference between multiband and single-band completeness. Multiband fitting recovers additional sources at faint magnitudes across a range of colors, particularly toward relatively red W1-W2 colors view at source ↗

**Figure 6.** Figure 6: Differential completeness as a function of Spitzer magnitude, defined as the fraction of Spitzer-COSMOS sources recovered in WISE per magnitude bin. Single-band (blue) and multiband (red) results are shown for Spitzer ch1 (left) and ch2 (right). Both methods achieve near-unity completeness at bright magnitudes, while multiband fitting remains complete to fainter limits. The largest improvement is observed … view at source ↗

**Figure 7.** Figure 7: Differential reliability (purity) as a function of WISE magnitude, defined as the fraction of WISE sources in each magnitude bin that have a Spitzer-COSMOS counterpart. Single-band (blue) and multiband (red) results are shown for W1 (left) and W2 (right). Multiband fitting maintains high reliability while extending to fainter magnitudes, with the most significant improvement observed in W2 demonstrating th… view at source ↗

read the original abstract

We present a new multiband extension to the crowdsource photometric pipeline, enabling simultaneous fitting across multiple imaging bands in crowded fields. The core idea is that multiple images of the same part of the sky should have the same sources at the same locations; only the fluxes in the different images should be allowed to vary in fitting. The framework also allows us to use all images of a given region to detect faint sources, with configurable weighting among the different bandpasses as appropriate for different source spectra. Similar concepts are already present in other crowded field packages like DAOPHOT and DOLPHOT; we now include it in the crowdsource fitting approach. We describe the mathematical formulation of the multiband fit and demonstrate its performance using the Wide-field Infrared Survey Explorer (WISE) W1 and W2 imaging as a concrete application. The multiband algorithm improves flux consistency and reduces band-to-band positional scatter relative to independent-band fitting. We test the method on unWISE coadded tiles spanning both sparse and crowded regions and quantify improvements in photometric agreement and astrometric stability. This framework provides a general foundation for future multiband crowded-field catalogs.

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.

Desk Editor's Note private letter to a colleague

Crowdsource now has a shared-position multiband mode that improves flux consistency on WISE data but the positional scatter reduction is automatic from the model.

read the letter

The paper implements a joint multiband fit inside crowdsource where sources share one position across bands and only fluxes change. They apply it to unWISE W1 and W2 tiles in both sparse and crowded regions, using all images for detection with optional spectral weighting. This is a direct extension of the existing single-band pipeline and matches ideas already in DAOPHOT and DOLPHOT, so the conceptual step is modest but the code-level integration is new for this tool. The practical payoff is a single catalog with internally consistent positions and fluxes, which matters for large survey processing. The flux-consistency gains look plausible if the shared-position assumption holds, and the paper reports tests on photometric agreement. The claimed drop in band-to-band positional scatter, however, is built into the model by design and does not show independent astrometric improvement. The shared-position premise is reasonable for most WISE sources but will fail for objects with real color-dependent centroids or motion; the paper should flag how often that happens. No free parameters or invented quantities appear in the description. This is useful incremental work for anyone building crowded-field catalogs from WISE or similar IR surveys. It is solid enough to warrant referee time so the implementation details and quantitative validation can be checked.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces a multiband extension to the crowdsource photometric pipeline for simultaneous fitting across imaging bands in crowded fields. The core model enforces identical source positions across bands while permitting band-specific fluxes to vary, and it incorporates multi-band data for source detection with configurable weighting. The approach is demonstrated on WISE W1 and W2 coadded tiles in both sparse and crowded regions, with the central claim being improved flux consistency and reduced band-to-band positional scatter relative to independent per-band fits.

Significance. If the flux-consistency results hold under the shared-position assumption, the method offers a practical addition to the crowdsource framework for producing more consistent multi-band catalogs in crowded fields. It extends established ideas from packages such as DAOPHOT and DOLPHOT with a configurable detection weighting scheme. The non-tautological component (flux agreement) could benefit analyses requiring cross-band photometry, provided the assumption of fixed positions is valid for the target population.

major comments (2)

[Abstract] Abstract: the claim that the multiband algorithm 'reduces band-to-band positional scatter relative to independent-band fitting' is a direct consequence of the shared-position constraint (core idea paragraph). Inter-band position differences are identically zero by model construction in the joint fit but can be nonzero in independent fits; this therefore supplies no independent evidence of improved astrometric performance.
[Demonstration on WISE data] Demonstration on WISE data: the abstract states that improvements in photometric agreement and astrometric stability are quantified, yet no specific numerical results (e.g., RMS flux differences, median positional offsets with uncertainties, or statistical tests) are supplied. Without these concrete metrics and baseline comparisons in the results section, the central performance claims cannot be evaluated.

minor comments (2)

The abstract references DAOPHOT and DOLPHOT but supplies no citations; adding the standard references would place the work in context.
[Mathematical formulation] Mathematical formulation section: the configurable weighting for multi-band detection is described but would benefit from an explicit equation or numerical example showing how weights are chosen for different source spectra.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful and constructive review. We address each major comment below and have revised the manuscript to incorporate the suggested clarifications and additions.

read point-by-point responses

Referee: [Abstract] Abstract: the claim that the multiband algorithm 'reduces band-to-band positional scatter relative to independent-band fitting' is a direct consequence of the shared-position constraint (core idea paragraph). Inter-band position differences are identically zero by model construction in the joint fit but can be nonzero in independent fits; this therefore supplies no independent evidence of improved astrometric performance.

Authors: We agree that the elimination of inter-band positional differences is a direct and intended consequence of the shared-position model rather than an independent demonstration of improved astrometric precision. The joint fit enforces consistency by construction, which we view as a practical advantage for producing unified multi-band catalogs. We will revise the abstract and the relevant discussion to remove any implication that this constitutes evidence of superior astrometry and instead emphasize it as a model feature that improves catalog consistency. revision: yes
Referee: [Demonstration on WISE data] Demonstration on WISE data: the abstract states that improvements in photometric agreement and astrometric stability are quantified, yet no specific numerical results (e.g., RMS flux differences, median positional offsets with uncertainties, or statistical tests) are supplied. Without these concrete metrics and baseline comparisons in the results section, the central performance claims cannot be evaluated.

Authors: We acknowledge that while the manuscript states that improvements are quantified, the results section does not present the specific numerical metrics and statistical comparisons with sufficient detail. In the revised manuscript we will add explicit values for RMS flux differences between bands, median and RMS positional offsets (with uncertainties), and direct comparisons to independent-band fits, including any relevant statistical tests, supported by additional tables or figures. revision: yes

Circularity Check

1 steps flagged

Band-to-band positional scatter reduction is by construction under the shared-position model

specific steps

self definitional [Abstract]
"The multiband algorithm improves flux consistency and reduces band-to-band positional scatter relative to independent-band fitting."

The framework defines that 'multiple images of the same part of the sky should have the same sources at the same locations; only the fluxes in the different images should be allowed to vary in fitting.' Band-to-band positional differences are therefore identically zero by model construction in the joint fit, while nonzero in independent-band fits. The asserted reduction is a direct consequence of the shared-position assumption rather than an independent result.

full rationale

The paper's core model enforces identical source positions across bands by definition, with only fluxes varying. The claimed reduction in band-to-band positional scatter therefore follows tautologically from this constraint when compared to independent fits (where positions may differ). Flux consistency improvements are not similarly forced, but the strongest claim bundles both, making the overall result partially circular. No other derivation steps reduce to inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review prevents exhaustive identification; the approach rests on standard crowded-field photometry assumptions plus the new multiband constraint.

axioms (1)

domain assumption Sources share identical positions across all bands
Explicitly stated as the core idea enabling joint fitting.

pith-pipeline@v0.9.0 · 5506 in / 1099 out tokens · 72609 ms · 2026-05-07T17:11:55.682335+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

19 extracted references · 17 canonical work pages

[1]

1999, XSPEC: An X-ray spectral fitting package,, Astrophysics Source Code Library, record ascl:9910.005 http://ascl.net/9910.005 Astropy Collaboration, Robitaille, T

Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068

work page doi:10.1051/0004-6361/201322068 2013
[2]

2016, DOLPHOT: Stellar photometry,, Astrophysics Source Code Library, record ascl:1608.013 http://ascl.net/1608.013

Dolphin, A. 2016, DOLPHOT: Stellar photometry,, Astrophysics Source Code Library, record ascl:1608.013 http://ascl.net/1608.013

2016
[3]

Dolphin, A. E. 2000, PASP, 112, 1383, doi: 10.1086/316630

work page doi:10.1086/316630 2000
[4]

S., Krolewski, A., Qu, F

Farren, G. S., Krolewski, A., Qu, F. J., et al. 2025, PhRvD, 111, 083516, doi: 10.1103/PhysRevD.111.083516

work page doi:10.1103/physrevd.111.083516 2025
[5]

F., & White, M

Krolewski, A., Ferraro, S., Schlafly, E. F., & White, M. 2020, JCAP, 2020, 047, doi: 10.1088/1475-7516/2020/05/047

work page doi:10.1088/1475-7516/2020/05/047 2020
[6]

Cosmological constraints from unWISE and Planck CMB lensing tomography

Krolewski, A., Ferraro, S., & White, M. 2021, JCAP, 2021, 028, doi: 10.1088/1475-7516/2021/12/028

work page doi:10.1088/1475-7516/2021/12/028 2021
[7]

2014, AJ, 147, 108, doi: 10.1088/0004-6256/147/5/108

Lang, D. 2014, AJ, 147, 108, doi: 10.1088/0004-6256/147/5/108

work page doi:10.1088/0004-6256/147/5/108 2014
[8]

W., & Schlegel, D

Lang, D., Hogg, D. W., & Schlegel, D. J. 2016, AJ, 151, 36, doi: 10.3847/0004-6256/151/2/36

work page doi:10.3847/0004-6256/151/2/36 2016
[9]

M., Lang, D., Schlafly, E

Meisner, A. M., Lang, D., Schlafly, E. F., & Schlegel, D. J. 2019, PASP, 131, 124504, doi: 10.1088/1538-3873/ab3df4

work page doi:10.1088/1538-3873/ab3df4 2019
[10]

M., Lang, D., & Schlegel, D

Meisner, A. M., Lang, D., & Schlegel, D. J. 2017, AJ, 154, 161, doi: 10.3847/1538-3881/aa894e

work page doi:10.3847/1538-3881/aa894e 2017
[11]

C., & Saunders, M

Paige, C. C., & Saunders, M. A. 1982, ACM Trans. Math. Software, 8, 43 14

1982
[12]

Portillo, S. K. N., Speagle, J. S., & Finkbeiner, D. P. 2020, AJ, 159, 165, doi: 10.3847/1538-3881/ab76ba

work page doi:10.3847/1538-3881/ab76ba 2020
[13]

, keywords =

Sanders, D. B., Salvato, M., Aussel, H., et al. 2007, ApJS, 172, 86, doi: 10.1086/517885

work page doi:10.1086/517885 2007
[14]

The Astrophysical Journal Supplement Series , author =

Schlafly, E. F., Meisner, A. M., & Green, G. M. 2019, ApJS, 240, 30, doi: 10.3847/1538-4365/aafbea

work page doi:10.3847/1538-4365/aafbea 2019
[15]

F., Green, G

Schlafly, E. F., Green, G. M., Lang, D., et al. 2018, ApJS, 234, 39, doi: 10.3847/1538-4365/aaa3e2

work page doi:10.3847/1538-4365/aaa3e2 2018
[16]

2007, ApJS, 172, 1, doi: 10.1086/516585

Scoville, N., Aussel, H., Brusa, M., et al. 2007, ApJS, 172, 1, doi: 10.1086/516585

work page doi:10.1086/516585 2007
[17]

Stetson, P. B. 1987, PASP, 99, 191, doi: 10.1086/131977

work page doi:10.1086/131977 1987
[18]

S., Connolly, A

Szalay, A. S., Connolly, A. J., & Szokoly, G. P. 1999, AJ, 117, 68, doi: 10.1086/300689

work page doi:10.1086/300689 1999
[19]

L., Eisenhardt, P

Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868, doi: 10.1088/0004-6256/140/6/1868

work page Pith review doi:10.1088/0004-6256/140/6/1868 2010