Astrophotography diagnosis of NGC 7380: Noise / underexposure, backfocus error and 1 other
MasterL30×15s31 oct. 2026
The Doc examined this image of NGC 7380 (master, L, 30×15s). Estimated overall technical quality: 4/10. 3 defects found: Noise / underexposure (severity 4/5), Backfocus error (severity 3/5), Focus miss (severity 3/5).
&w=1920&q=75)
Info
- Cible
- NGC 7380
- Date
- 31 oct. 2026, 23:38
- Lune
- Croissant croissant 40.9% (21.8° d'alt., 92.1° de la cible)
- Site
- Bortle 4 · transition rural/banlieue (VIIRS)
- Position
- 22h47m22s · +58°08'21"
Conditions were rather favourable. Bortle 4 gives a healthy sky background and the background map confirms it: plane gradient of only 1 %, R2=0.09, no oriented light pollution signature. The Moon is at 40.9 % illumination and 21.8 deg altitude, but 92.1 deg from the target: at that separation its influence on NGC 7380 is negligible, and no lunar gradient is measured. In other words nothing in the sky explains the weak signal: it comes entirely from the exposure. With a sky like this you can afford much longer luminance subs without fear of swamping the background, and that is exactly what is needed here.
Setup
- Type d'image
- Master
- Télescope
- Celestron C11 2800@F\10.0
- Caméra
- PlayerOne CCD Poseidon-M PRO
- Filtre
- L
- Exposition
- 30×15s
- Phase de lune
- Gibbeuse décroissante (65 %)
- FOV
- 45.8'
The C11 at f/10 is poorly matched to NGC 7380. The field is 0.76 deg wide while the cluster and its nebulosity span nearly 25 arcmin, so the target fills the frame with no margin and most of the nebular structure loses its context. Above all, sampling at 0.44"/px is heavily oversampled: the measured 6.7 px central FWHM equals about 2.9 arcsec, meaning you spread the signal over 6 to 7 times more pixels than needed. On settings, gain 125 and -10 deg C are healthy, while offset 30 leaves 99 % dark clipping driven by 15 s subs that are far too short at f/10. A 0.63x reducer, or 2x2 binning, would bring both field and sampling back into a usable range.
The diagnosis in detail
Three observations hang together, and they share one root: the imaging train and the acquisition plan are not matched to this target. The most damaging is the lack of signal. With 30 subs of 15 s the sky background stays 2.38 ADU from the black point and 99 % of the distribution is clipped in the shadows, meaning you are mostly digitizing read noise. NGC 7380's nebulosity shows only as a few wisps around the central cluster, while a Bortle 4 sky left you all the headroom you needed for much longer exposures.
Optically the measurement is clear: the center is round (elong 1.10 against a 1.08 floor) but FWHM goes from 6.7 px at center to 9.3 and 10.6 px at the bottom of the field, a corner/center ratio of 1.35 spread all around the frame with no diagonal asymmetry. That is field curvature / backfocus error, not tilt. The TL zone shows an isolated 1.54 elongation, but its fit is disturbed by a close neighbour and I do not treat it as a discriminant.
Finally the central plateau itself at 2.9 arcsec indicates focus was not optimal; a thermally unsettled C11 drifts quickly. Tracking, on the other hand, is good: PA dispersion of 33.7 deg and no coherent elongation direction, the mount is not at fault.
Priority actions
- Lengthen subs to 60-120 s and accumulate several hours instead of 7.5 minutes, this is what changes everything here
- Correct the imaging train backfocus in 0.5 to 1 mm steps, then recheck corner/center FWHM on a test run
- Refocus with a Bahtinov mask after 45 to 60 minutes of thermal settling, with regular automatic refocus
- Bin 2x2 or fit a 0.63x reducer to bring sampling toward 0.9"/px and widen the field on NGC 7380
- Raise the offset slightly to remove the 99 % dark clipping
