Manual 2 star align?

[bilbo]

Dear friends,

I use a Celestron NexStar SE8 scope with an ASI120MC camera. Unfortunately due to the Alt-Az mount I am only able to shoot 10-20-30 sec pictures which occasionally have too few stars on them (for example in the vicinity of M51).
I cannot use automatic stacking on those pictures (or am only inexperienced , not even automatic 2-star align because of the low quality details.

I tried manual align, selecting the same star on every image and it was succesful - but as expected, the align procedure didn't adjusted the different rotation of the pictures... So the stacking was done but the details far from the reference star were 'smeared' along a section of a circle.

Do you now how to help this issue? Is there any way to do 2 star aligmnent in manual mode?

Thanks a lot in advance for the advices!

Have a good day and clear skies!

Krisztian
from Hungary

[fabdev]

Hello,
manual alignment works only with one star. By the way, please post or send by email the first and last image of a sequence (even in JPEG) maybe there is a setting to force the automatic alignment.
Fabio.

[bilbo]

Dear Fabio,

you are kind I attached 2 light frames (with the maximum offset between them), a dark frame and the screenshot of the result of the manual align (you can see the little arcs in the lower right corner due to field rotation).

It's most likely my fault as I am still learning the tricks of processing deepsky images The source images may be wrong too. There are 18 light frames, the offset between them varies in distance and direction too. All of the auto align methods give ghost images (I left all options on the default).

Thanks for helping!

Krisztian

PS: do you think a 2-star manual align would be useful? as having only the 1 star method in the manual mode, there is no way to deal with field rotation (contrary to the 'Star pattern translation+rotation' auto align method)?

[fabdev]

Hello,
yes, the "Two stars" method will work (increase the Radius to 100), but not perfectly, because the images also have a difference in scale. Maybe they were taken with a zoom lens which moved? cause the temperature?

So, the correct alignment should be done with "Star pattern" but it's not possible on the JPEGs, there are too many defects. If you like send me by email the original RAW FITS image #01 and #34. I think that with some hot pixel correction inside the RGB Demosaic the "Star pattern" could detect the required stars.

[bilbo]

Dear Fabio,

you are extremely kind to offer that but I wouldn't like to steal your time If you have a minute and check the pictures, I would be very happy but only if you don't have better thing to do I am searching for a solution because with my hardware setup, I will not be able to shoot much better pictures and if I cannot find a method to stack them, I don't know what to do...

Here are the files: https://drive.google.com/drive/folders/ ... sp=sharing

BTW: increasing the star size worked with Star Pattern, the process ended succesfully - but the result was very strange. I made a screenshot from my preprocessing settings and the result picture. What do you think I made wrong? (This time not only 2 but all of the light frames were in the list).


I have never ever thought that there can be a difference in the scale too, I learn something new every day (not a joke, really, I have started astrophotography since a few months) . Strange, although I used an f/6 focal reductor, this session was almost the last that night so the optical tube+accessories were out for hours...

Any little advice from you would be heavily appreciated - and thanks for taking care of the issue of a stranger

Best:
Krisztian

PS: again, manual 2-star align wouldn't be the optimal solution for such inferior light frames? It would handle the translation, rotation and scale too and would be fool-proof

[fabdev]

Hello Krisztian,
I think that we can solve, but I don't understand: the files on your Google Drive are coloured BMP (?)
Don't you have the original FITS RAW files? Which software did you use to download them?
Greetings,
Fabio.

[bilbo]

Dear Fabio,

I used FireCapture and only used BMP output format because I thought it is not compressed just as FITS but I can open and edit .bmp files with common image processors too (like Gimp or IrfanView). Seems I made a wrong decision...?

Have a beautiful day!!

Krisztian

[fabdev]

Hello, working with BMP has at least 3 problems:

1) The dynamic is only 8 bit, with FITS you would get 12/16 bits.

2) The dark frame cannot be corrected well. For example, in your sequence try to open an image, for example #14, then open a dark frame, then subtract the dark frame from the image and tune the visualization: as you can see the correction is not good at all.

3) The RGB demosaic in BMP is now fixed (and it's low quality, "blocky 2x2", see wth "Edit pixel"), but the dark frame should be corrected before demosaic, so again the dark frame cannot be corrected well.

I suggest to take another sequence, and save in FITS (without RGB debayer!). Then align, demosaic and stack the images inside an astro program. Later you may save the result in BMP to make the final fixes with your preferred graphic program,
Fabio.

[bilbo]

Dear Fabio,

thanks for your valuable suggestions! I will do as you said!

Just out of curiosity: why isn't debayering in the capture program preferred?

Best:
Krisztian

[fabdev]

Hello, here are 4 reasons:

1) The dark frame must be corrected before debayering.

2) Debayering is a filter, and you may need to change the algorithm (the one used in your BMPs was not good, look at the shape of the stars).

3) In case of a discovery, you must have the original images.

4) Storing debayered images takes more space on your harddrive or backup.

[bilbo]

Dear Fabio,

sorry for the delay and many thanks for your kind advices!! I am sure I have to learn a lot more!

I think I will re-shot that sequence as soon as there will be a clear sky - can't wait to try it again!

Have a good day!
Krisztian