A fixed extinction ratio and grey dimming with constant red light both fit 31 ZTF colour pairs of ASASSN-24fw under stated errors

We test two explanations for the small colour change in the 2024–2025 eclipse of ASASSN-24fw. We use 31 public ZTF DR24 g/r pairs and 141 baseline pairs. We propagate the linear colour correction and the two-band covariance. We fit errors in both bands and include a shared baseline constraint. A fixed effective extinction ratio gives k = 1.02750 and chi-square 25.18 for 30 degrees of freedom. Grey variable light plus constant red light gives a flux-flux slope A = 1.002692 and chi-square 22.96 for the same degrees of freedom. Both models fit under the stated errors. Eleven calibration and data-selection checks retain both fits. In the constant-light model, nonnegative g flux requires constant r flux of at least 0.2685% of baseline r flux at the best fit. Many constant colours give the same line, so the source of this light is not identified. The best-fit curves differ most by 0.039 mag at an r fade of 2.49 mag. None of the event pairs has an r fade between 1 and 3 mag. This gap limits the test of the curve shape. The result applies to these two bands and these error assumptions. It does not test the published spectra, polarization, or full dust models. We preserve the data, provenance, paired measurements, code, bootstrap samples, figures, and reproduction instructions.

ASASSN-24fw · ZTF DR24 · stellar eclipses · constant light · colour degeneracy · dust extinction · flux-flux relation · public data