Star Alignment on Lights Only Halpha, SII and OIII

sandconp

Well-known member
I have six light frames (2) Hapha, (2) SII and (2) OIII and I don't need flats or darks since the images are already calibrated. BPP and WBBP don't work because PIxinsight says it needs at least 3 light frames for it to work. I generally have lots of light, darks and flats when I image with my own telescope. But I acquired these images from Telescope LIve by scheduling some time for NGC2264 and I was able to download these six frames and I want to process them.

What would be the first process to process these images?, Would Star Alignment be the first step and what would be the second step? I am pretty familiar with PI but I am little bit green when it comes to stacking and aligning because I am spoiled with using the scripts.
 
Last edited:
Yes, once you have calibrated images StarAlignment followed by ImageIntegration.
One small point, sometimes when receiving calibrated data there may still be hot pixels running around that require CosmeticCorrection.
Too few frames or lack of dithered data will make this more important. However, I am familiar with TelescopeLive data- and hot pixels are generally not an issue (they did a good job keeping current and high ca quality calibration frames).
-adam
 
So I added my six frames (2 of each Halpha, OIII and SII) to Star Alignment and I pointed to the first Halpha frame as the reference frame and ran the Star Alignment process.

I then got 6 output files with a suffix of cal-R.xisf.

03-00-52_NGC2264_Halpha_T-30_300s_cal_r.xisf (example)

I have never run ImageIntegration before and I am not sure what that process does?

And what actually does the Star Alignment process do because I get the same amount of output files and my input only suffixed with xisf?
 
SA "does what it says on the tin" - it finds matching stars between the reference frame and the other frames, and produces modified versions of the frames such that the stars are aligned (i.e. at the same image-relative (x,y) position). The images may be shifted and/or rotated and/or scaled (and - if requested - distortion corrected) to align the stars. You now need to separately integrate your aligned filter image pairs, producing three integrated images, one for each filter (since ImageIntegration requires a minimum of three frames you will need to add each of your images twice - this will do no harm). Then you will need to combine the separate (monochrome) images to produce an RGB colour image (e.g. using ChannelCombination).
 
So I completed Star Alignment and Image Integration and then moved onto to Pixel Math and I am getting some strange looking stars. They look liked they are multi-color stars. I also tried LRGB combination and had the same problem. Before I did the Pixel Math, I checked the OII, SII and Halpha frames and all of the stars lined up so I believe that something in Pixel Math or LRGB combination is causing these multi-color stars.
 

Attachments

  • NGC2264-2.small.png
    NGC2264-2.small.png
    305.8 KB · Views: 48
If you could upload your three integrated filter images (to some cloud site) and post a readable link I could have a look at them.
 
It looks as thought your two Oiii frames are aligned, but the frame pairs for Sii and Ha are not quite aligned. Could you upload the separate frames.
 
I just upload the six original fits files that I downloaded from Telescope LIve. There are (2) of each HA, SII and OIII.

My steps were the Star Alignment, Image Integration and I also had to re-add one frame one of each to get by the restriction of needing at least 3 frames. I did not do the Cosmic Correction because I wasn't sure when to do that. I then proceeded to my LRGB combination or Pixel Math and that's where I noticed the problem with the stars.

Thanks for your help.
 
I aligned all frames to the first Oiii frame (StarAlignment with default settings), producing a set of 6 fully aligned frames.
I then loaded each pair of frames twice into ImageIntegration (giving 4 frames), and integrated with default settings.
I then used ChannelCombination to combine the filters in SHO format. A simple unlinked STF then gives this:
1611505938785.jpeg

All the stars seem well-aligned. One or two artefacts that could be rejected with a few more frames, but a nice image.
 
As you can see, I still have the multi-colored stars when I zoom in after I did the channel combination after following your new instructions.
 

Attachments

  • PI.NGC2264.Forum2.small.png
    PI.NGC2264.Forum2.small.png
    257.2 KB · Views: 41
You can see here that I do not get this problem. Are you sure you aligned all images (in a single run of StarAlignment with all six frames), and that you integrated and combined the registered images (_r suffix) output by SA?
1611560221530.png
 
Last edited:
Here are my exacts steps that I performed.

1. I took the (6) raw Fit files from Telescope Live 2 HA, 2 OIII and 2 SII and put them into Star Alignment choosing one of the frames as reference image. I am assuming that it doesn't make any difference which fit file I choose as the reference image? I then landed up with (6) frames in my Star Aligned Folder with a suffix of cal.r.xisf (2) per filter of HA, SII and OIII.

2. I then ran the ImageIntegration process (3) separate times for each for HA, OIII and SII and also re-added the same aligned frames from the Star Alignment for each filter for a total of (4) frames per filter under the Add Files tab.

3. Out of the ImageIntegration process, PixInsight created a combined frame for each filter of HA, SII and OII for a total of (3) frames and I continued on to the next process of ChannelCombination.

I am obviously doing something wrong.

Please Advise if my process looks correct.

Thanks so much.
 
That process looks like exactly what I did. If you could upload your .r files and your single filter integrations to your dropbox folder, perhaps I can try to identify at what point the difference is appearing.
 
OK I deleted my old NGC2264 Dropbox Folder and created a new one.

There are six star aligned fits file with a suffix of r.xisf for each filter.

I then saved the HA, OII and SII fits file that were created from ImageIntegration. Also in the folder is the Sho.Fit file from ChannelCombination that shows the multi-colored stars.


Let me know if you need anything and thank you so much for you help.
 
I believe your "HA integrated" image is an integration of the unregistered Ha frames. If you recalculate this integrated frame, making sure you use the registered HA images, I think you will be OK.
 
I am not sure what you mean and how can you tell the Ha Frames are not registered and how do I fix it?

What did I do wrong and in which step(s)?
 
he means you may have accidentally loaded the non-_r versions of the Ha exposures into ImageIntegration when you ran that step.
 
how can you tell the Ha Frames are not registered
I use the Blink process. if you load the three integrated images (HA, OIII, SII) and blink them slowly, you can se the small "jump" between the HA image and the other two (easier to see if you define a "zoomed in" preview in the blink window).
When you look at the HA integration, the stars are "doubled", indicating that it was integrated from unaligned images.
If you blink the HA integration and the two registered HA files, it is clear that the HA integration is not aligned, while the two registered images are.
QED :)
 
Back
Top