Astrophotograph of Comet C/2019 Y4 (ATLAS)

This is my astrophotograph of the comet C/2019 Y4 (ATLAS), which I observed for 7.5 hours on the night of the 26–27 March 2020. I took 1194× 20-second frames in total (animated in the video below). Due to the building cloud and shifting focus as the telescope tube cooled down, I only aligned and stacked the best 808 of these frames (the ones that would register in DeepSkyStacker). This final astrophotograph therefore has a total exposure time of 4.5 hr.

As you can see from the video, the comet moves relatively fast against the background of stars. The integrated image can therefore either be based on frames aligned on the moving stars, or based on frames aligned on the moving comet. Aligning on stars and averaging pixels leaves a streak from the comet, or aligning on the comet and averaging pixels leaves star streaks (as in the image below).

To make an integrated image without streaks from either the comet or the stars (as in the first astrophotograph above), I used a special type of averaging technique called ‘kappa-sigma (κ-σ) clipping’ to make one image that rejects the comet while tracking the stars, and to make another image that rejects the stars while tracking the comet: .

Actually, κ-σ clipping is a fairly standard integration mode for astrophotography, and one that I use for almost every astrophotograph I make. It’s a useful technique to reject deviant pixels (for example, rejecting aeroplane streaks). It takes two parameters: the number of iterations and the standard deviation multiplier used (κ). For each iteration, the mean and standard deviation (σ) of each pixel in the stack are computed. Each pixel value in the stack that is more than κ×σ is rejected. The mean and standard deviation of each of the remaining pixels in the stack are then computed again in the next iteration, and so on.

When aligning on the comet and integrating using a small enough value of κ, the majority of pixel values in the stack at every location away from the comet have the sky background colour. Therefore star pixel values in the stack tend to be greater than κ×σ meaning they get rejected from the final stack average and leave just the comet behind. Vice versa, when aligning on the stars and integrating using a small enough value of κ, the comet gets rejected. There is a trade-off to make as small κ leads to more pixel values being rejected and more noise but also more effective rejection of the stars or comet. I find a κ value in the range 1.2 to 1.5 works well.

Once I had isolated the comet and stars in separate images, I blended them together in Photoshop to create an image where both the stars and the comet are sharp.

Finally, based on this, a note to explain the naming convention, which was set by the International Astronomy Union (IAU) in 2003: Comets are designated by the year of their discovery followed by a letter indicating the half-month of the discovery (starting with A for the first half of January, excluding I and Z, and going up to Y for the second half of December) and a number indicating the order of discovery. Prefixes are also added to indicate the nature of the comet, with P/ indicating a periodic comet, C/ indicating a non-periodic comet, X/ indicating a comet for which no reliable orbit could be calculated, etc.

C/2019 Y4 (ATLAS) is therefore a non-periodic comet, which was the fourth comet discovered in the second half of December 2019. It was discovered by the Asteroid Terrestrial-impact Last Alert System (ATLAS).


  • 808× 20-s light frames
  • 60× dark frames
  • 60× flat frames
  • 54× dark flat frames


  • Explore Scientific ED 102 mm Apo f/7 refractor
  • Sky-Watcher EQ5 PRO SynScan GOTO equatorial mount
  • Altair Hypercam 294C PRO colour fan-cooled camera
  • Altair 60mm guide scope
  • Altair GPCAM2 AR0130 mono guide camera


  • Sharpcap
  • PHD2
  • DeepSkyStacker
  • Adobe Photoshop CC 2019

Leave a comment