2

Stacking by Sequator 100 dng photos of night sky taken with a smartphone resulted in a (beautiful, I would say) populated star field where undesired arc streaks are seen all over the field. It is like as the program treated the vast majority of stars as wanted, while (perhaps the less brighter) others are just rendered as they moved along the 100 frames period.

Align stars was chosen then accumulation.

Question: Is this a common problem using Sequator, or at least happened to some of you? Is there any trivial setting or way to redo the stacking? Thanks.

Attached is a compressed copy of the final photo. 

enter image description here

the above is cross posted in Photography SE

https://photo.stackexchange.com/questions/124748/arcs-in-stacked-star-field

Mike G
  • 18,640
  • 1
  • 26
  • 65
Alchimista
  • 893
  • 6
  • 9
  • 3
    Cross-posting is generally frowned upon, as you may receive answers on both sites and users may not be able to find other good answers related to it. – WarpPrime May 06 '21 at 12:40
  • 2
    I wonder if those are just "hot pixels" that don't move on the sensor like the real stars do, so that the stacking spread them out in an arc. I don't think the software processes each star separately, they can't be "skipped stars" mingled with "recognized stars". Notice that they are all probably 1 pixel wide, whereas stars are much wider. – uhoh May 06 '21 at 13:00
  • @fasterthanlight I am aware therefore I have signalled the occurrence. I am also thinking of joining the more relevant material by future editing. – Alchimista May 06 '21 at 13:13

1 Answers1

4

Those are probably just "hot pixels" that don't move on the sensor like the real stars do, so that the stacking spread them out in an arc.

I don't think the software processes and treats each star separately, they can't be "skipped stars" mingled with "recognized stars". Notice that they are all 1 pixel wide, whereas stars are much wider.

uhoh
  • 31,151
  • 9
  • 89
  • 293
  • 2
    Not thinking of a "deliberate" separate treatment but I was thinking of somehow different sets. But your idea of hot spots is quite sensible. As I understand the software could make arcs and not straight strikes because (relative to) of the arc motion that it fixes. Is this that you mean? – Alchimista May 06 '21 at 13:11
  • 1
    @Alchimista yes that's just what I mean; the stars move in arcs that are all centered on the north or south celestial poles. Stacking rotates the images around a point to make them line up, and so the fixed hot pixels end up rotating instead. Most cell phone software masks out the hot pixels in the sensor when it makes JPEGs for us, but these are .dng (raw data) images so the hot pixel artifacts are clearly visible. If you look at a single .dng very closely and carefully you should be able to see them. – uhoh May 06 '21 at 14:06
  • 2
    I must consider a most careful use of dark frames and base etc... I think yours is the explanation... @uhoh – Alchimista May 06 '21 at 14:41
  • @Alchimista I'll recommend somehow closing the cross-posted duplicate in Photography SE to avoid answer-fragmentation. – uhoh May 06 '21 at 14:45
  • see my comment to fasterthanlight. I think I can manage this, at least for a while. – Alchimista May 06 '21 at 14:52
  • I noticed that, giving that the camera was pointing about South and latitude was about 47 N, the arcs have the wrong curvature. This seems to me a proof that the points were indeed not moving and sequator did a good job. Does this simply geometrical reasoning proof or corroborate your answer? They should indeed acquire that curvature as opposed to that of the apparent motion of stars, when the latter is removed. – Alchimista May 07 '21 at 08:15
  • hang on, I'm trying to use Astronmetry.net to figure out where this is exacly... – uhoh May 07 '21 at 12:59
  • aim was Berenice, which is indeed there in the right half of the photo middle height. Latitude perhaps 43 N not 47. But the geometrical reasoning seems linear (although I don't know the details of Sequator algorithm, it make sense that if it fixes something moving along a concave arc, it makes a convex arc of something immobile, or viceversa....). I am really convinced of your answer. – Alchimista May 07 '21 at 13:29
  • @Alchimista I see, How much time is there between the first and last exposure? – uhoh May 07 '21 at 14:40
  • 1
    100 frame 2 sec each separated by 1 sec. About 300 if the apps worked smoothly. But I think your answer is correct ( even if it would be another reason :)) because it makes much sense. – Alchimista May 07 '21 at 15:26
  • unrelated to the Q/A. I've tried astronometry but it failed telling me about objects in the south emisphere while it was a clear pic of Orion :)) – Alchimista May 07 '21 at 16:58