Astrophotography diagnosis of NGC 6924: Background chroma noise and noise / underexposure
Processed
The Doc examined this image of NGC 6924 (processed). Estimated overall technical quality: 7/10. 2 defects found: Background chroma noise (severity 2/5), Noise / underexposure (severity 2/5).
&w=1920&q=75)
Info
- Cible
- NGC 6924
- Site
- Bortle 5 · banlieue résidentielle (VIIRS)
A moonless night (0%) under a Bortle 5 sky: that is the right window for a broadband galaxy, since the absent Moon rules out any lunar gradient and Bortle 5 remains manageable on a target as bright as NGC 6946. The background measurement confirms it: only 2% plane gradient, flat radial profile, no notable residual light pollution after processing. The limit comes from Bortle 5 itself, which caps SNR on the faint spiral arm extensions and the outer HII regions. To go further on this target you need integration time, not more gradient correction.
Setup
- Type d'image
- Traitée
- Télescope
- Celestron c8
- Caméra
- Zwo 1600
- Phase de lune
- New moon (0 %)
C8 plus ASI1600, roughly 0.6 arcsec/px at 2032 mm focal length: you are oversampled for typical seeing, which the measured 5.5 px FWHM confirms, about 3.3 arcsec. The target sits comfortably in the field with margin all around, framing is clean. Two options: binning 2x2 at capture or in processing would give back SNR without losing real detail, since resolution is atmosphere-limited rather than optics-limited. Otherwise, the uniform PSF across all nine zones points to well set backfocus and collimation on this C8, which is never a given.
The diagnosis in detail
Optically and mechanically there is nothing to flag. The PSF panel shows nine nearly identical zones, FWHM between 5.41 and 5.79 px, elongation peaking at 1.05 at center against a field floor of 1.02, corner to center ratio at 1.011 and horizontal and vertical asymmetries below 0.02. In other words, no sensor tilt, no backfocus error, no residual coma and no signature of drift or guiding oscillation. On a C8 that is a clean result validating both collimation and the focusing chain.
The real subject is signal to noise. The background to black point margin of 1.9 shows you are not in the sky-limited regime, and it reads directly in the image: the background carries colored mottling in clumps of a few tens of pixels, and the outer spiral arms plus the small pinkish HII regions barely stand out from the grain. Processing was clearly restrained, with no abusive denoising, which is a good call, but it cannot make up for missing signal.
The logical next step is acquisition: more hours, color channels first, and longer subs to lift the background. The rest of the chain is already up to standard.
Priority actions
- Add integration time, color channels first, to push back the background chroma mottling
- Lengthen sub exposures to reach the sky-limited regime (target e_margin between 5 and 10)
- Apply targeted chroma noise reduction on the background, leaving luminance untouched
- Consider 2x2 binning of the chrominance, consistent with the 0.6 arcsec/px oversampling
