Plate solve fails on NGC 7023
Stacking program, astrometric (plate) solver, fits file viewer
Brought to you by:
han59
Menu
▾
▴
I have an AGOPTICAL iDK 14.5 (2464 fl) and QHY268M setup at Howling Coyote Remote Observatory (HCRO) and only have the D80 database installed. I am imaging NGC 7023 and ASTAP fails to plate solve the target inside NINA. ASTAP has worked well for every other target except M 2 where I had the same issue.
If I load one of the 2 minute subs directly into ASTAP it still fails to solve.
Any help would be appreciated.
Hi Jeffery,
Can you share the image involved for testing? You could uploaded it to nova.astrometry.net and share the link or by Google drive. You could also attach to your reply and I will remove it as soon I have downloaded it
cs, Han
Han,
Hopefully this is what you need?
https://nova.astrometry.net/status/16225854
Thanks,
Jeff
Hi Jeff,
The site nova.astrometry.net has not solved you upload. That is very rare, so there must be something special with this image. Unfortunately I can not retrieve the .fits file from the nova.astrometry.net because it was not solved. I can only retrieve a jpeg.
Luckily I can solve the jpeg. Image size is 0.55 x 0.36 degrees. It look like it contains a lot of hot pixels. That could be the cause of the problem
I like to have a look to the original FIT image. Can you attach it to your next posting? Or share with something else. E.g upload to ufile.io and share the link
cs, Han
Han,
Yeah, I have only had luck with Platesolve3 being able to solve it.
I can't seem to be able to upload the file directly here (it says the file is too big) so hopefully this link will work. Just copy the file and do what you wish with it just don't save it here.
https://drive.google.com/file/d/1fJKiSE1yHB50g27lmlJ-rJabTm3y36hd/view?usp=drive_link
Thanks,
Jeff
I need permission for viewing the file. See request.
cs, Han
Hi Jeffery,
The hot pixels are not a problem. The image solves if you force the binning to 4. The problem with long focal lengths is that the stars are spread out over many pixels and binning helps to improve the SNR and detection.
So I would suggest to force binning always to 4 in Nina (ASTAP solver settings in Nina). Tell me if this helps with solving.
cs, Han
Han,
Appreciate the help! If I just choose binning 4 in the plate solve section of "options" in NINA is that sufficient? And I assume that does not change my actual images saved? So plate solving mid-sequence to check for drift will do the binning after the image is saved as BIN 1?
CS,
Jeff
No luck. I still can't get ASTAP to solve any of my NGC 7023 raw FITS images.
Can you load the image in ASTAP and compare the settings with the attached screenshot? If different try the red marked settings. Then it should should solve without problems.
Setting ASTAP binning to 4 does not effect your bin 1x1 image. It only instruct ASTAP to bin 4x4 the image provided in memory only. The image at your hard disk will not be effected
The very odd thing is if I open sigma the FOV height is 0.37 and the solve fails. If after failing I change it to 0.36 and re-run on the same open image it fails again. But if I open the image and change the FOV height to 0.36 before running, it immediately solves. But each time I open sigma, the FOV height is 0.37 regardless if I saved 0.36.
Last edit: Jeffery Richards 2 days ago
Yes a little weird. When you enter a small field-of-view or longer focal length it works reliable.
I like to dig a little deeper to find out why this happens because I have seen it before and thought it was solved. This typically happens for long focal lengths. Can you share all your NGC7023 images and some other object images series for testing?
For the moment, I would increase the focal length in Nina to 110% since it will likely give a more reliable solving.
cs, Han
Han,
Here is a folder with a night's worth of NGC 7023 that doesn't solve and another night's worth of Sh2-136 that does solve.
https://drive.google.com/drive/folders/1zQKugUueuokGhNQASfGF7MJM6klbeWJ1?usp=sharing
Thanks,
Jeff
Hi Jeff,
You have to give me permission again to view the file(s)
One thing which could explain the problem is that the images are taken in blue. The stars in the Gaia database have a magnitude which is compatible with clear view, not with a blue filter. The blue filter will make the reddish stars look fainter. Looking to the B-V of some stars, I suppose they are reddish.
For astrometric solving, what would help in any case is to take the images without a filter (clear view or an uv/ir block filter). This will also require less exposure time then with a blue filter.
cs, Han
Hi Han,
My initial plate solve is always done with the Luminance filter and 7023 still failed. I didn't happen to take Luminance images for this target (yet) so all I have is the RGB files in the folder. Of course that doesn't help when I have a plate solve going on to verify image drift during a session as it will just plate solve the last image while taking the next one so that has to work with whatever filter is active.
I think I fixed your file access.
Thanks,
Jeff
The strange thing is that nova.astrometry.net is also failing to solve. I uploaded four images and they all fail to solve. That is very abnormal.
But tweaking the FOV to a smaller size helps ASTAP is an indication. The same star count for a smaller FOV will force a higher amount of stars extraction from the database and then there is a solver match.
For solving the star density extracted from the image should match reasonable to the star database. But the field-of-view of the star database is square and different.
Looking the a wider NGC7023 image:
https://skyandtelescope.org/online-gallery/ngc-7023-the-iris-nebula-4/
is showing that NGC7023 is locally a star poor area like a dark nebula. Outside the image there are many more stars. I assume this unbalance is causing the solver to fail when not tweaked.
So I have no solution except the tweak with setting focal-length at 110% but it seems a phenomena related to NGC7023 only and your setup field-of-view.
cs, Han
Han,
I do appreciate the effort and really have no idea what the issue is. As I mentioned, I had this exact same issue trying to image M 2. It simply would never plate solve to the point that I gave up and moved on to another target (this was before I switched to PS3 as my blind solver). I do not have any of those failed images for M 2 so I can't share them.
I'm reluctant to change my fl number because I like my FITS info to be as accurate as possible. By having PS3 as my blind solver I think I'll be okay and can live with the rare failed solve warnings that pop up.
Thanks again,
Jeff