Astrophotography diagnosis of M33: Backfocus error and clipped stars
Processed
The Doc examined this image of M33 (processed). Estimated overall technical quality: 8/10. 2 defects found: Backfocus error (severity 2/5), Clipped stars (severity 2/5).
&w=1920&q=75)
Info
- Cible
- M33
- Position
- 1h33m49s · +30°38'36"
A 27% waning crescent is only a mild nuisance, even for M33 which is a broadband target hungry for SNR. At that phase the Moon rises late and its glow stays localized: if you shot early in the night its effect on the background is negligible, which the clean, gradient-free background of this image confirms. For this kind of galaxy the real limiting factor is transparency and local light pollution, not those 27%. If you want to push the faint outer disk, still favor moonless nights.
Setup
- Type d'image
- Traitée
- Phase de lune
- Dernier croissant (27 %)
- FOV
- 1.01°
The 1.01 degree wide field suits M33 well, its disk filling about 70% of the frame width: the galaxy breathes without being clipped, with room left for the outer extensions. The measured central FWHM of 5.6 px points to generous sampling (rather oversampled for the effective seeing), so you have headroom for software 2x2 binning to gain SNR without losing real detail. The monochrome camera and the strongly marked red HII regions suggest an L-RGB composite with an Ha layer, which is the right call on this target.
The diagnosis in detail
The main diagnosis is optical and measured, not eyeballed: the PSF goes from 5.6 px at center to 6.5-7.1 px in the corners, with the center at minimum and a symmetric rise across all four angles. That is the classic signature of field curvature or a backfocus off tolerance by a few tenths of a millimeter. Elongation stays low everywhere (1.05 to 1.17, field floor at 1.05) and PA dispersion is 21.5 degrees, which rules out a tracking problem: the stars are round, just broader toward the edges.
On the processing side the bulge core keeps its yellow-cream gradient with no flat plateau, which is healthy. Several bright field stars, however, show pure white cores with no chromatic gradient, the classic loss of B-V information during the stretch. The four-branch spikes visible on those same stars are synthetic, added in post, and are an aesthetic choice rather than a defect.
The sky background is clean, with no instrumental gradient measurable here since the galaxy fills the frame, and the HII regions come out a clear pink-red, a sign of correct color calibration.
Priority actions
- Adjust the sensor-to-corrector spacing in 0.5 mm steps and re-check the four corner FWHM values
- Protect stellar highlights during the stretch (luminance mask or piecewise histogram transfer) to preserve bright star color
- Consider software 2x2 binning to exploit the oversampling and gain SNR on the faint extensions
