Horsehead Nebula

What to improve next time

• warn4 of 17 subs were dropped by grading…33_R_10.00_180.00s_0000: few stars: 694 < 1776 (cloud/haze?), elongated: ecc 0.62 > 0.6; …46_R_10.00_180.00s_0002: soft: FWHM 4.10 > 4.06 px; …09_R_10.00_180.00s_0004: soft: FWHM 4.34 > 4.06 px, few stars: 553 < 1776 (cloud/haze?), elongated: ecc 0.62 > 0.6; …01_R_10.00_180.00s_0005: few stars: 900 < 1776 (cloud/haze?). If a sub looks fine to you, keep it with 'skystack grade --keep-all' or loosen the limits.
• warnStars are soft: FWHM 5.8" (3.2 px)Typical seeing gives 2-4". The most common cause is focus - use a Bahtinov mask or autofocus, and refocus as the temperature drops. Poor seeing or tube currents are the other suspects.
• warnStars are elongated (elongation 0.53; 0 = round)Usually tracking: guiding off or struggling, polar alignment, wind, or balance. If only the corners are elongated it is optics (tilt or the flattener spacing) instead.
• warnNo flatsVignetting and dust shadows are left in the image. 20-30 flats (same focus and camera angle as the lights) fix both. 'skystack background' can hide smooth vignetting, but not dust donuts.
• warnNo darksHot pixels and amp glow stay in. Take 20-30 darks at the lights' exposure, gain and temperature.
• tipSharpness changed during the session (2.3 to 3.9 px)Focus drift (temperature) or changing seeing. Refocusing every hour or after a 1-2 C temperature drop keeps every sub as sharp as the best one.
• tipThe field rotated 0.11 degrees during the sessionField rotation on an equatorial mount means polar alignment is off.
• tipTransparency changed: some subs were 65% dimmerHaze or the target getting lower. It is corrected, and weighting favours the clearer subs.
• tipOne sub had far more rejected pixels: …9-29_02-46-36_R_9.90_180.00s_0001Usually a satellite, plane or passing cloud. Rejection removed it - see the rejection map.
• tipTotal exposure 39 min: more time is the biggest improvementNoise falls with the square root of the total time: 4x the time halves the noise and brings out much fainter detail.
• goodAlignment adapted to field distortionSome subs needed an affine fit (typical for wide fields low in the sky, where atmospheric refraction squeezes the field); all aligned cleanly.
• goodNoise reduction at the ideal: 11.6 effective subs of 13Weighting and outlier rejection wasted nothing.

1. Calibration

Lights

17 x 180 s

Darks

none

Bias

0

Flats

none

Read noise

-

Dark signal

-

Hot pixels repaired

0

Sensor temperature

9.9 to 10.0 C

• No darks or bias: lights are NOT calibrated (hot pixels and offset stay in).
• No flats: vignetting and dust shadows are not corrected.

2. Sub grading

3. Alignment

4. Integration

Output

masters\masterLight_HorseheadNebula_13x180s_R.fits

Rejection

winsorized sigma clip, low 4.0 / high 3.0

Values rejected

0.453% low, 1.389% high

Weights

noise

Cropped to full coverage

2982 x 2910 (lost 4.1%)

Dither spread

86.1 px

Random noise (measured from differences, so stars and nebula cancel)

Per sub