Astrophotography diagnosis of M16: Bad seeing and clipped stars
ProcessedSHO3x30x600s
The Doc examined this image of M16 (processed, SHO, 3x30x600s). Estimated overall technical quality: 8/10. 2 defects found: Bad seeing (severity 3/5), Clipped stars (severity 2/5).
&w=1920&q=75)
Info
- Cible
- M16
- Site
- Bortle 5 · banlieue résidentielle (VIIRS)
- Position
- 18h18m50s · -13°51'49"
Conditions were favorable for this target. A 25 percent waning crescent contributes almost nothing to the background, and in SHO on an emission nebula like M16 the narrow passband neutralizes most of a Bortle 5 sky anyway. It shows in the result: the background is neutral, with no urban gradient or lunar tint, and the faint outer nebulosity comes through without fighting a milky sky. With 15 cumulative hours under these conditions the signal to noise achieved is consistent; the limit here is not sky brightness but sky stability, turbulence is what capped the resolution.
Setup
- Type d'image
- Traitée
- Télescope
- Planewave CDK 12,5"
- Caméra
- Moravian C4 monochrome
- Filtre
- SHO
- Monture
- Paramount MX+
- Exposition
- 3x30x600s
- Phase de lune
- Dernier croissant (25 %)
- FOV
- 48.8'
The setup is perfectly matched to the target. A 0.81 degree field places M16 and its nebular envelope centrally with comfortable margins, without cutting the extensions, and the 12.5 inch CDK focal length delivers the resolution wanted on the Pillars of Creation, which are the real subject here. The Paramount MX plus does its job: no measurable elongation, PA spread with no drift signature, so nothing to fault in tracking on 600 s subs. SHO is the right choice under Bortle 5 on an HII region. The only mismatch is that this combination samples finer than the seeing of those nights could deliver, hence stars broader than the optics can actually produce.
The diagnosis in detail
The image is of a good standard and the measurements confirm it on two key axes. Tracking is exemplary: elongation of 1.01 to 1.09 depending on zone, right at the field floor, with no preferred direction or curvature. On 600 s subs, that is the signature of a well-tuned mount and healthy guiding. On field optics, the corner to center FWHM ratio is 0.857, so the corners are actually tighter than the center: there is no edge degradation, and the automatic tilt pre-verdict is not supported by the panel, where no readable diagonal asymmetry appears. That candidate is therefore rejected.
The real limiting factor is PSF width. Stars are round but large, with a wide spread between zones, from 7.9 px at bottom right to 13.5 px along the bottom edge. Such variation across so small a field cannot be explained by the optics; it points instead to stacking subsets acquired under differing turbulence, the stack averaging uneven quality. The direct consequence is visible on the bright stars, whose spread energy clips more readily into flat white cores at stretch.
The processing itself is well controlled: a coherent and pleasing SHO palette, a clean background with no residual gradient or notable chroma noise, soft transitions in the faint nebulosity with no plastic look. The gains ahead are therefore upstream, in sub selection, rather than in post-processing.
Priority actions
- Sort subs by FWHM with SubframeSelector and reject or down-weight the softest frames before integration
- Add a short-exposure set per filter and blend stellar cores in HDR to remove the flat whites
- Automate temperature-step refocusing in NINA to stabilize FWHM through the night
- Apply gentle deconvolution on the linear image to tighten PSFs without generating dark halos
