Astrophotography diagnosis of M16: Backfocus error and over-denoising
Processed3 filtres Baader 6.5nm : S, H et O75x300s
The Doc examined this image of M16 (processed, 3 filtres Baader 6.5nm : S, H et O, 75x300s). Estimated overall technical quality: 8/10. 2 defects found: Backfocus error (severity 2/5), Over-denoising (severity 2/5).
&w=1920&q=75)
Info
- Cible
- M16
- Position
- 18h18m48s · -13°48'00"
You imaged under an 85 to 100% moon, which is very harsh for broadband but almost irrelevant here thanks to the Baader 6.5 nm S, H and O filters: the full moon does not attack an emission target in narrowband. The background stays clean with no detectable lunar gradient, fully validating the narrowband choice in these conditions. The trade-off is lower SNR on OIII and SII (the target emits less in those lines), hence the need for denoising. For an identical future session, favor extending OIII/SII time rather than shortening it, the moon will not penalize you in Ha.
Setup
- Type d'image
- Traitée
- Télescope
- SkyWatcher 150 PDS Quattro
- Caméra
- ZWO ASI585MM Pro
- Filtre
- 3 filtres Baader 6.5nm : S, H et O
- Monture
- Sky-Watcher EQ6-R Pro
- Exposition
- 75x300s
- Phase de lune
- de 85% à 100%
- Notes
- Traitement PixInsight, BlurXTerminator, NoiseXTerminator, StarXTerminator
- FOV
- 36.0'
The Quattro 150 PDS (600 mm focal) + ASI585MM Pro combo yields a 0.60 deg field that frames M16 tightly and centered on the pillars, a sensible compositional choice that highlights the central structure. Sampling is fine (small 585 pixels), explaining the perfectly fine ~3.5 px FWHM under the night's seeing. The point to watch remains the Newtonian's coma corrector backfocus: a 0.5 mm step adjustment would bring the corner/center ratio under 1.15 and tighten the corners. The Baader S/H/O 6.5 nm trio is ideally suited to this emission nebula.
The diagnosis in detail
On the PSF plate the center C is round (elong 1.02, at the field floor) and PA dispersion reaches 31 deg, ruling out a tracking elongation: stars are pointlike and guiding is excellent. The only clear pattern is a homogeneous symmetric rise of FWHM toward the four corners (ratio 1.20, TL 4.22 px versus C 3.49 px), with no radial comet tail or diagonal asymmetry: signature of a mild backfocus / field curvature issue of the Quattro, not tilt nor coma. BlurXTerminator already tightened the corners well, the final visual impact is minor.
On processing, the BlurX + StarX + NoiseX combination is well controlled, the pillars retain excellent detail. The OIII background in the right half however betrays slightly heavy denoising (smoothed, grainless look), a logical consequence of limited SNR in that line under full moon and moderate Bortle. It is not damaging at this scale but it is the most direct lever for improvement.
Overall a clean image, perfect tracking, calibration with no visible donuts or walking noise, balanced SHO palette. The room for improvement is purely instrumental (backfocus) and acquisition (more OIII/SII signal).
Priority actions
- Adjust the coma corrector backfocus in 0.5 mm steps to bring the corner/center FWHM ratio under 1.15
- Increase OIII and SII integration time to raise SNR and ease denoising
- Moderate NoiseXTerminator strength and apply it via a background-only mask
