AMSP-Only 2 of 3 nights merged

Software Siril
1

Hi all, pretty new user here, so forgive my lack of terminology. This may not be an AMSP problem, but it seems like the right place to start.

Just getting into doing multinight sessions. I’m using an ASIAir Mini on my SD102 F5 scope. I’m using the method described in post 2 here to align subsequent nights with a sub from the original session, and in mid-May did a two night session using AMSP. Really easy!

I did 2 nights on M101 at the end of May that also merged great using AMSP. However, I tried to add a third night, done June 16 using the same method of manually aligning the camera in the ASIAir Mini as for the other nights, to the lights from the first 2 nights. The end result was nights 1 & 2 merged and night 3 as a separate file.

Happy to provide whatever image, log, etc. files that will help to narrow down the issue. Thanks in advance.

2

Let me be a bit more complete …

  • I am manually aligning the camera with a sub from an older session.
  • The night 3 image is shifted about 15-20 degrees “northwest” of the merged night1-night2 image, but when I looks at sub samples from the three nights, they look closely oriented.
3

It the third night is not merged that means the OBJCT keyword is not the same.

4

Thanks @lock042 , I’ll have a look at that. and compare. I have your excellent book, between that & this forum, I should be able to resolve that.

1 Like
5

Unfortunately, the OBJECT entry in the FITS header is set to ‘M101’ in the third night, just as it is in nights 1 & 2. Any other suggestions? I can try to run AMSP again or do it manually following the procedure in your book, but of course I’d prefer not to if there’s no need.

6

So the exposure time must be different.

7

Same for all 3 nights - 300 sec.

8

Please paste me filename of three different stacking results

9

Here they are. They were randomly chosen from the lights from each night in question.

  • Light_M 101_300.0s_Bin1_533MC_gain100_20260525-032327_-10.0C_0030

  • Light_M 101_300.0s_Bin1_533MC_gain100_20260527-012856_-9.9C_0013

  • Light_M101_300.0s_Bin1_533MC_gain100_20260616-041538_-10.0C_0041

  • List item

10

Apologies - those are the lights filenames! I’ll post the stack names shortly.

11

Here they are. The first name is of the stack of two nights, the second is the third night.

  • M_101_FILTER_77x300sec_G100_O70_T-10C_2026-05-24
  • M101_FILTER_41x300sec_G100_O70_T-10C_2026-06-15

I’ve also included the FITS headers from the two stacks:
FITSHeaders_20260525_20260616.zip (5.2 KB)

12

This may be normal for multinight sessions, but I notice the SITELAT/LONG and OBJECTRA/DECs are different in the FITS headers of the two stacks.

  • 05-24 stack:

    • SITELAT 42.0788
    • SITELONG 3.09606
    • OBJRA 14 03 13.596
    • OBJDEC +54 19 4.377

  • 06-15 stack:

    • SITELAT 42.0789
    • SITELONG 3.09614
    • OBJRA 14 02 57 316
    • OBJDEC +54 21 10.975
13

So there’s a difference in the OBJECT name. M 101 versus M101

14

Agg, I’m sorry not to have caught that! I’ll change the name in the lights to something consistent, and it should merge OK.

Is this something the Asiair did/does? I certainly don’t remember doing it, I platesolved off a sub and went from there. Anyway, another good piece of knowledge, many thanks!

15

@lock042 just to let you know the rename of the OBJECT attribute worked, and I added a fourth night to the integration for about 14 hours of integration time. Still working on the postprocessing, but I’m really pleased with the quality from having multiple nights data…