Astrophotography diagnosis of ngc281: Clipped stars and over-denoising
ProcessedL-pro23x300s
The Doc examined this image of ngc281 (processed, L-pro, 23x300s). Estimated overall technical quality: 7/10. 2 defects found: Clipped stars (severity 2/5), Over-denoising (severity 2/5).
&w=1920&q=75)
Info
- Cible
- ngc281
- Position
- 0h53m09s · +56°37'27"
Moon at 0%, so no lunar contribution to the sky background: conditions were optimal in that respect, and the weak histogram margin (e_margin 0.86) reflects a background sitting close to the black point after processing rather than moonlight. On an emission target like NGC 281 the L-Pro is a reasonable choice under moderately polluted skies, but it passes a lot of continuum: on a dark moonless night like this one, a dual-band could have been considered to boost Ha contrast. Dark clipping at 0.02% is negligible, the offset is fine.
Setup
- Type d'image
- Traitée
- Télescope
- C8 réducteur 6,3
- Caméra
- ZWO ASI294MC Pro
- Filtre
- L-pro
- Monture
- AM3
- Exposition
- 23x300s
- Phase de lune
- 0%
- FOV
- 50.3'
The C8 reduced to f/6.3 gives a 0.84 degree field, framing NGC 281 tightly but coherently: the nebula fills the frame without being cut, a valid composition. The 2.7 px central FWHM indicates correct sampling and focus, with no penalizing oversampling. The corner-to-center degradation (2.7 to 3.5 px) stays low for a reduced SCT, meaning the reducer backfocus is well set, often a sore point on this configuration. The AM3 holds 300s with no measurable elongation (1.04 at center). The real lever is not the gear but the time: 23x300s, under two hours, is thin for this target at this focal length.
The diagnosis in detail
On the acquisition side, nothing to complain about. The aberration inspector gives a center at 2.74/2.62 px FWHM for a 1.04 elongation, meaning round stars at the measurement floor. Corners rise to 3.3-3.55 px with elongations of 1.09 to 1.18, a corner/center ratio near 1.3 which, for a Schmidt-Cassegrain with a 6.3 reducer, reflects the normal residual field curvature of that combination rather than tilt: there is no clear diagonal asymmetry between opposite corners, all four degrade comparably. The AM3 tracking over 300s is faultless, with scattered PAs, a sign of round stars rather than a common elongation direction.
The two retained defects are therefore entirely post-processing matters. The bright star cores are clipped to pure white, erasing stellar colors while the DSS confirms they exist: that is a stretch pushed without highlight protection. In parallel the sky background shows no grain, too smooth for 1h55 of cumulative exposure, the signature of somewhat heavy denoising compensating for missing signal.
The common cause sits right there: integration time. With twice the subs you would need neither that much smoothing nor that forced a stretch, and the faint Pacman extensions would gain markedly in finesse.
Priority actions
- At least double the integration time (aim for 4 to 6h on NGC 281) to stop compensating missing signal with smoothing
- Redo the stretch with a mask protecting star cores to recover stellar colors
- Lower the denoise strength and restrict it to the background with a luminance mask
- Consider a dual-band filter on this emission target to gain Ha contrast at the same focal length
