flipped images not registering?

Questions, support and discussions
buzz
Posts: 6
Joined: 08 Nov 2019, 18:31
Contact:

flipped images not registering?

Post by buzz » 08 Nov 2019, 18:44

Hi - this is my first post to this forum. I am an experienced imager with a long background in image processing and I was trying out ASTAP 29 Oct beta on some QHY OSC flats, darks, biases and lights.

The final stack has a problem- it has a distinct vertical boundary with the joining of two images. https://www.dropbox.com/s/tpvgb6lxoney8 ... t.jpg?dl=0

On closer inspection and comparing with my PixInsight stack of the same files, I realised that the image is made of 2/3 of the original image and 1/3 of an inverted image from the other end. I did have meridian flips in my imaging sequence and I think the registration process has got confused somewhere along the line.

Have I done something wrong, maybe with the registration settings, or is this a gremlin?

han.k
Posts: 288
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: flipped images not registering?

Post by han.k » 08 Nov 2019, 21:33

Hello Buzz,

A flipped or rotated image or both is not a problem. It looks like to match routine made an error. The image is pretty large and full of well focused stars. I can astrometric (plate) solve it only if I increase the setting "Maximum number of stars to use" to 1200. Once solved the star annotation shows the left part is wrong compared with the main image as you suggested.

I would suggest to change the setting as in the blue rectangle "Tetrahedron detection". The other settings could be used for solving but are not required for stacking.
Untitled.png
Untitled.png (17.66 KiB) Viewed 190 times
If it doesn't help, could provide someinput files for testing?

Also, what exposure time did you use? A shorter exposure time will probably make stacking easier but could be less desirable.

The star colour also looks not coming from the ASTAP AstroC debayering algorithm. All the star centers are white due to saturation. Are the input files raw or processed/debayered in the camera or by PI? Having not the raw files is a handicap for stacking. Here an example of a colour image out of ASTAP https://www.vwsnoorddrenthe.nl/phpBB/vi ... f=13&t=165

p.s this in the HNSKY forum. A few lines below is the ASTAP forum. I can't move message to other forum. Post there the next topic.

Han

buzz
Posts: 6
Joined: 08 Nov 2019, 18:31
Contact:

Re: flipped images not registering?

Post by buzz » 09 Nov 2019, 18:38

Hi Han, many thanks for the input. I will give it another shot.
These are 1200-second exposures with a duo-narrowband filter. Yes, some stars have saturated cores but these are replaced by the color from shorter exposures taken through an LP filter. (I confirmed the files were ok by running them through PixInsight and DSS without issue)

I am really excited about your program. I'm currently finishing up a beginner's book on astrophotography to compliment my others. I was looking for a cross-platform version of DSS to recommend and I think I found it.

Sorry about posting in the wrong window....

Post Reply