Astrophotography diagnosis of SH2-126: Sensor tilt
Rawl.ultimate300s13 sept. 2026
The Doc examined this image of SH2-126 (raw, l.ultimate, 300s). Estimated overall technical quality: 6/10. 1 defect found: Sensor tilt (severity 3/5).
&w=1920&q=75)
Info
- Cible
- SH2-126
- Date
- 13 sept. 2026, 22:26
- Lune
- Gibbeuse décroissante 91.2% (sous l'horizon)
- Site
- Bortle 3 · rural peu pollué (saisi)
- Position
- 22h34m52s · +38°39'43"
Very favorable conditions. The 91% waning gibbous Moon was below the horizon during the exposure, so no lunar contribution is possible: the background map confirms it, with a near-zero oriented component (2% amplitude, R2=0.01). The stated Bortle 3 shows in a dark, clean sky background with no measurable light-pollution gradient. The reported 2.5 arcsec seeing has no real effect here: at 5.75 arcsec/px you are heavily undersampled and turbulence vanishes inside the pixel. In short, the night limits nothing in this frame; everything degrading it comes from the hardware and the integration time, not the sky.
Setup
- Type d'image
- Brut
- Télescope
- SIGMA135-ASI6200MC
- Caméra
- ZWO ASI6200MC Pro
- Filtre
- l.ultimate
- Monture
- azeq6
- Exposition
- 300s
- Phase de lune
- Premier croissant (5 %)
- Seeing
- 2.5
- FOV
- 15.57°
The setup matches the target well. SH2-126 is a very large, faint nebula in Lacerta, and the 15.6 deg field of the Sigma 135 on the ASI6200MC frames it comfortably, with the arc structure nicely centered. Sampling at 5.75 arcsec/px is very coarse, but that is the normal regime for a wide-field rig: you lose stellar detail and gain flux, the right trade-off on a low surface brightness target. The L-Ultimate is consistent with an emission target. On settings, gain 101 (unity), offset 50 with negligible dark clipping (0.02%) and sensor at -10C are all healthy. The low background margin (2.2) is normal for dual-band under Bortle 3, not a sign of underexposure to fix.
The diagnosis in detail
The diagnosis comes down to one clear instrumental defect plus a limitation inherent to a single exposure. The Moffat measurements show a fully exploited center (FWHM 2.0/1.5 px, elong 1.33 against an OSC floor of 1.14) and a degradation that is neither radial nor symmetric: the top-left corner blows up to a 4.0 px major axis with 2.06 elongation, while the right edge stays at 1.14 elongation and the bottom-left at 2.1 px. This single-axis signature, with a 1.62 corner/center ratio and 0.264 vertical asymmetry, is that of a tilted focal plane, not a backfocus error which would hit all four corners equally.
At f/2.2 the depth of focus is a few tens of microns: a few hundredths of a millimeter of adapter or sensor tilt is enough to produce this spread. So it is mechanical and adjustable, not an inherent optical limit of the lens.
The second point is noise, but it should be read for what it is: an isolated dual-band sub on a very faint target under a dark sky necessarily yields a grainy background and a barely emerging nebula. That is not a settings error, it is a lack of cumulative time. Tracking itself is clean: no coherent elongation direction across the field, the mount is not at fault.
Priority actions
- Fix the imaging train tilt first: cross-tighten everything, then iterate with a tilt adjuster aiming for under 15% FWHM spread between corners
- Shoot a test star field after each adjustment iteration to confirm the four corners converge
- Build up integration: target 6 to 10 hours on SH2-126 to lift the nebulosity out of the noise
- Enable dithering between subs to avoid correlated noise at stacking
