Astrophotography diagnosis of V?ga: Sensor tilt and clipped stars
RawIr-Cut5s07 sept. 2026
The Doc examined this image of V?ga (raw, Ir-Cut, 5s). Estimated overall technical quality: 5/10. 2 defects found: Sensor tilt (severity 3/5), Clipped stars (severity 2/5).
&w=1920&q=75)
Info
- Cible
- V?ga
- Date
- 07 sept. 2026, 18:45
- Lune
- Gibbeuse croissante 85.9% (sous l'horizon)
- Site
- Bortle 6 · banlieue dense (VIIRS)
- Position
- 18h36m59s · +38°48'03"
Decent conditions for this test. The Moon, although 85.9 percent gibbous, is below the horizon: no lunar pollution is possible, and the measured background confirms it (azimuthal anisotropy 0.01, plane gradient of barely 5 percent with an R squared of 0.15, so not significant). Bortle 6 remains penalising for faint deep sky, but on a magnitude 0 star in a 5 s exposure it has no effect at all. The stated 2.5 arcsecond seeing explains most of the central FWHM: at 0.44 arcsecond per pixel, 3.5 px is about 1.5 arcsecond, even better than the declared seeing, a sign that the short exposure partly freezes the turbulence.
Setup
- Type d'image
- Brut
- Télescope
- C11
- Caméra
- ZWO ASI533MC Pro
- Filtre
- Ir-Cut
- Monture
- EQ6r pro
- Exposition
- 5s
- Phase de lune
- Dernier croissant (17 %)
- Seeing
- 2.5
- Notes
- F=6,3
- FOV
- 23.2'
The setup matches the goal: a C11 at f/6.3 on an EQ6-R, a square 3008 px ASI533MC sensor, a 0.39 degree field and 0.44 arcsecond per pixel sampling. That is heavily oversampled for 2.5 arcsecond seeing, fine for planetary work or a PSF test like this one, but unsuited to deep sky: a shorter reducer or 2x2 binning would make the signal far more efficient. On settings, gain 100 and offset 20 are healthy, but 10.3 percent dark clipping with a background to black point margin of 2.7 confirms underexposure: normal at 5 s, to be fixed with longer exposures for deep sky. Finally the 22.4 degree sensor temperature shows cooling is off, and should be enabled (minus 10 degrees) to limit dark current.
The diagnosis in detail
The defining issue in this image is tilt. The field center is sharp with a major axis FWHM of 3.46 px and an elongation of 1.17, close to the OSC floor, so optics and tracking do their job correctly on axis. The corner to center FWHM ratio however reaches 1.63 and the vertical asymmetry peaks at 0.334: the top edge (TL 5.77 px, T 4.81 px, TR 5.09 px) is clearly more bloated than the bottom edge (BL 2.95 px, B 3.94 px, BR 4.14 px). This single axis degradation with a preserved center is the mechanical signature of a tilted focal plane, not of a backfocus error, which would have hit all four corners equally.
The rest is more about shooting context than genuine defects. Vega being magnitude 0, its core is fully clipped and its scattering halo dominates the center of the frame: unavoidable here, but it rules out any colour measurement on it. The green background cast is the normal behaviour of a debayered OSC sensor without white balance and will vanish with colour calibration. The 29.4 degree PA dispersion and the absence of coherent elongation confirm there is no tracking problem over 5 seconds, as expected from an EQ6-R.
Finally, the measured underexposure (margin 2.7, 10.3 percent dark clipping) and the sensor at 22.4 degrees are not defects on a test sub, but would be disqualifying for a deep sky acquisition.
Priority actions
- Fix the imaging train tilt: retighten crosswise, check adapter flatness, then run a 90 degree sensor rotation test to locate the source
- Enable camera cooling (minus 10 degrees) before any session, the sensor is at 22.4 degrees here
- Lengthen the sub exposure substantially for deep sky, the 2.7 background to black point margin indicating clear underexposure
- Consider 2x2 binning or a shorter reducer: 0.44 arcsecond per pixel is heavily oversampled for 2.5 arcsecond seeing
- Check the cleanliness of the Schmidt corrector plate and the Ir-Cut filter, since the scatter around Vega is asymmetric
