What am I doing wrong?
Stacking program, astrometric (plate) solver, fits file viewer
Brought to you by:
han59
Menu
▾
▴
When I first started using astap I got good results. But for the last year or so, almost all my results look like this... (M1 version on Mac OS 13.6.9)
I have the impression that the frame selected as reference image is one where the tracking was lost. So the stars are streaks instead of points. Inspect the image marked with a crown visually.
So load the images in the image tab, press analyse button and double click on the image marked with a crown to visually inspect.
Han
It automatically selects what it determines to be the best image. This is not a one-time problem, but is occurring with every data set.
Can you share a data set for testing?
Link to data set:
............
Last edit: han.k 2024-08-25
I have the images. The link in your I have removed now.
After debayering a single image i noticed that the colours are a little weird. Did you use a filter?
No filter. IR cut.
I'm in doubt about the debayer pattern. I noted that ASTAP select as a reference image a frame with all stars double. See attached.
Not what I see...
The problem is that ASTAP aligns on the hot pixels. Change the setting of "Ignore stars less then HFD" from 0.8 to 1.5 to filter them out. See attached screenshot.
The hot pixels form coloured streaks. Better to make and use darks.
The debayer pattern look okay.
If you have any other problem, just ask.
Cheers, Han
And check the image marked with the crown after starting the stacking.
I stacked your image with method average, so that will NOT remove hotpixels. Use for stacking method "sigma clip" to remove them.
I am using 29 darks. I manually removed all the images in which I saw star doubling. Change ignore to 1.5. I use Sigma Clip Average. But I still got the same result. No improvement.
BTW, both DWARF and Siril are correctly stacking these images.
Set the downsample from 0 to 2.
Change also the setting of "Ignore stars less then HFD" to 3.
Setting Ignore to 2 fixed it. Trying downsample now.
Thanks.
Downsample 2 is the best fix. You stars are not so sharp.
Keep "Ignore stars less then HFD" at 2.