Astrophotography diagnosis of NGC 6604: Clipped stars and background chroma noise
ProcessedAntlia ALP-T Dualband 3nm Ha & OIII 2" + Antlia ALP-T Dual Band 5nm SII&H-beta Filter 2"64×300s + 12×300s07 juin 2026
The Doc examined this image of NGC 6604 (processed, Antlia ALP-T Dualband 3nm Ha & OIII 2" + Antlia ALP-T Dual Band 5nm SII&H-beta Filter 2", 64×300s + 12×300s). Estimated overall technical quality: 6/10. 2 defects found: Clipped stars (severity 2/5), Background chroma noise (severity 2/5).
&w=1920&q=75)
Info
- Cible
- NGC 6604
- Date
- 07 juin 2026, 00:00
- Site
- Bortle 5 · banlieue résidentielle (saisi)
Bortle 5 with a 59% Moon: average conditions, but largely offset by the 3nm/5nm dualband on an emission target like the NGC 6604 region. The narrowband rejects most light pollution and moonlight glow, which explains a usable background despite the Moon. The background margin stays low (e_margin 1.06), normal for narrowband at 300s but leaving little background signal: on a moonless night you would gain depth on the faint nebular extensions.
Setup
- Type d'image
- Traitée
- Télescope
- ZWO FF130
- Caméra
- ZWO ASI2600MC Pro
- Filtre
- Antlia ALP-T Dualband 3nm Ha & OIII 2" + Antlia ALP-T Dual Band 5nm SII&H-beta Filter 2"
- Monture
- WarpAstron WD-20
- Exposition
- 64×300s + 12×300s
- Phase de lune
- 59%
- Notes
- Photoshop, PixInsight
Coherent, well-sized setup. At 1000mm f/7.7 on the ASI2600 you sample around 0.8"/px, so oversampled: your measured FWHM (~2.8-3.6px, ~2.3" at center) are very good and the WD-20 mount tracks cleanly (no directional elongation, scattered PAs). The field frames a rich Ha portion of Serpens around NGC 6604/Sh2-54, nicely filled. The Antlia dualband is the right choice here. Given the oversampling, a slight x2 binning in processing would boost SNR with no real detail loss.
The diagnosis in detail
Star shape is healthy across the whole field. The Moffat metrics give a central elongation of 1.09 against a 1.08 field floor, and a corner/center FWHM ratio of only 1.14, below the threshold that would indicate field curvature. The automatic backfocus pre-verdict (56% confidence) rests mainly on the horizontal asymmetry (0.23) driven by the L zone (FWHM 4.51): the PSF panel shows this zone is contaminated by merged bright neighboring stars that skew the fit, not by a real optical defect. So I do not retain backfocus-error. Tracking, collimation and backfocus are clean.
The real points are in processing. Many bright star cores are clipped to pure white, the result of a heavy global stretch without highlight protection. And the sky background, barely lifted off the black point (e_margin 1.06), lets chroma graininess rise during the stretch.
Finally, the rendering is very strongly red Ha-dominated; the OIII layer (and SII/Hbeta) seems underrepresented, which flattens the palette. Better narrowband balancing would add nuance and depth without distorting the target.
Priority actions
- Redo the stretch by separating the stars (StarXTerminator) to recover clipped cores and protect the highlights
- Reduce background chroma noise (SCNR/chrominance) and limit stretching of the underexposed background
- Rebalance the narrowband channels (raise OIII/SII) to break the red dominance
- Extend integration on a moonless night to gain background SNR on the faint nebulosity
