Star Rotation - Out of alignment

Jkulin

Well-known member
Apr 14, 2017
93
9
Hi All, I'm afraid I need to call upon your help again.

I have been working on NGC1491 and have about 18+ hours on it, but... I didn't realise that I had knocked my camera by about 11 degrees, some part half way through my captures.

I have tried stacking and it fails saying it

*** Error:

*** Error: RANSAC: Unable to find a valid set of star pair matches.

So does this mean that I now need to throw the whole data away and start again or can I do something to fix the rotation, here's a screen grab of the Ha which did stack: -
1609977275884.png
And here's just one of the unstacked images that won't stack/process: -
1609977376826.png
It would be a bummer if I did have to throw away the data, as I previously had made the same mistake on 5 hours of data on the same image.

Problem is that it is a bit tight in my observatory and just a simple knock can move it out without my realising.

Any thoughts or help would be appreciated.

Thank-you.

Regards,

John
 

pfile

PTeam Member
Nov 23, 2009
6,170
180
likely the problem are all those hot pixels. if SA detects them as stars then it can really mess up the star matching process.

in theory your use case is very easy for StarAlignment.

are you calibrating these frames? i would expect there to be far fewer hot pixels in a calibrated sub.

rob
 

Jkulin

Well-known member
Apr 14, 2017
93
9
likely the problem are all those hot pixels. if SA detects them as stars then it can really mess up the star matching process.

in theory your use case is very easy for StarAlignment.

are you calibrating these frames? i would expect there to be far fewer hot pixels in a calibrated sub.

rob
Hi Rob,

Thanks for your help.

Sorry for the slow reply, I have been snowed under with work.

I attempted to process with the WBPP and that is where it failed, it created a Master for the Ha but could not create for the OIII, I tried processing Manually and that failed as well.

I've come to the conclusion that they are so far off that It has prompted me to recommence the complete run of data.
 

fredvanner

Well-known member
Apr 17, 2019
828
90
70
Wells, Somerset, UK
I'm fairly sure that the problem is nothing to do with small offsets in alignment. It is all those hot pixels; have you explored the full capability of SA noise / hot pixel rejection? Could you upload full resolution images for three or four frames that fail to align?
 

Geoff

Well-known member
Mar 6, 2020
73
26
If the worst happens and you can't align the two sets, then stack each set separately and then align the two stacked images with DynamicAlignment. This is a manual process where you can pick corresponding stars in the two images.