Plate solving issues - Eurolad19

Hardware and software setups, autoguiding and plate solving parameters, etc..
Post Reply
Hamza Touhami
Site Admin
Posts: 564
Joined: Sat May 20, 2017 6:05 pm
Location: Phoenix,AZ
Contact:

Sat Dec 29, 2018 7:10 pm

The biggest problem so far I encounter is the plate solving. I loaded one of my photos (Veil nebula taken with a C14) and was unable to plate solve it. I tried different approaches. First I tried Astrometry.Net and despite I have the API key I always get the message: Invalid Key or Field error. When I go to Astrometry's website directly and plate solve my photo directly with them I got it solved, no problem.

When I use the option "single photo", change the focal length and pixel size to the size the photo was taken and use the RA/Dec which I got from the direct solve of Astrometry's website, I cannot solve it either, it gives me an error note with 5 different reason why there could be an error.

I tried also the UCAC catalog which I have on my computer, which use for PlatSolve2, I set the path to that location on my computer and the program cannot find it, then I copied the whole catalog into Prism directly and set the path via Settings Start catalog Path settings to the proper files. The program still cannot find the catalog.

So I am not sure how to get this Plate solving resolved. Kindly advise what else to try.

from Peter
Hamza Touhami | Site Administrator
Hyperion-Astronomy.com
eurolad19
Posts: 8
Joined: Mon Nov 13, 2017 9:08 am

Sun Dec 30, 2018 9:32 am

I try to upload a screenshot for you to see what I get
Hamza Touhami
Site Admin
Posts: 564
Joined: Sat May 20, 2017 6:05 pm
Location: Phoenix,AZ
Contact:

Mon Dec 31, 2018 9:11 am

you can try something like Dropbox or Onedrive and share th link here. No matter, let's address your questions:
1. PRISM does provide a few clues as to why a plate solve failed. Remember here: There is a localized and allsky plate solve engines in PRISM.
2. I am interested as to why the ASTROMETRY.NET site rejected your key through PRISM but not directly, did you set astrometry.net to "Online" in the settings?
3. The UCAC version you used does not seem to be the right one, please use the downloads section on our site to get the right catalog, UCAC4 works very well for the majority of focal lengths and applications.
Hamza Touhami | Site Administrator
Hyperion-Astronomy.com
eurolad19
Posts: 8
Joined: Mon Nov 13, 2017 9:08 am

Mon Dec 31, 2018 10:00 am

Screenshot (31).png
Screenshot (30).png
to your first point. I wanted to avoid to have another huge file download for the UCAC catalog as I assume from the tropdown menu in PRISM it accept all versions of this catalog. Anyway, to your second point re Astrometry.Net, yes I have set to "remote". I have attached two screenshots which you hopefully can see.
Hamza Touhami
Site Admin
Posts: 564
Joined: Sat May 20, 2017 6:05 pm
Location: Phoenix,AZ
Contact:

Mon Dec 31, 2018 10:18 am

your astrometry.net is pointed at the wrong URL, it should be http://nova.astrometry.net . The address you have now will only work if you have download the local engine with ALL its files.
Hamza Touhami | Site Administrator
Hyperion-Astronomy.com
eurolad19
Posts: 8
Joined: Mon Nov 13, 2017 9:08 am

Wed Jan 02, 2019 12:04 pm

Happy New Year to you Hamza

Thanks for your reply. Unfortunately I have no idea where I would have to change the URL to the one you suggested, i.e. http://nova.astrometry.net. Could you please kindly advise me where exactly I have to put this address in? Thanks.

On your advise I downloaded the UCAC_4 catalog via your website and installed it in Prism. I tried to plate solve this photo again with the new down;oad and still got a failed plate solve. Please see the attached two screenshots which I got trying to plate solve it.
Screenshot (37).png
Screenshot (36).png
Fellside
Posts: 11
Joined: Fri May 11, 2018 12:31 pm

Wed Jan 02, 2019 3:13 pm

Peter

Have a look at your first screen shot you posted.

Software Settings
Astrometry.NET
Third box down Solver URL
Remove what you have in the box.
Paste http://nova.astrometry.net

Hopefully that should get you sorted.

Graham
Hamza Touhami
Site Admin
Posts: 564
Joined: Sat May 20, 2017 6:05 pm
Location: Phoenix,AZ
Contact:

Wed Jan 02, 2019 10:20 pm

Thanks Graham, that's exactly correct.
Hamza Touhami | Site Administrator
Hyperion-Astronomy.com
Hamza Touhami
Site Admin
Posts: 564
Joined: Sat May 20, 2017 6:05 pm
Location: Phoenix,AZ
Contact:

Wed Jan 02, 2019 10:21 pm

Peter, if you succeed at uploading those images somewhere accessible, I'd be glad to take a look on my machine. For those interested, I will start a new series of video tutorials and I will address one simple subject in less then 2 minutes.
Hamza Touhami | Site Administrator
Hyperion-Astronomy.com
eurolad19
Posts: 8
Joined: Mon Nov 13, 2017 9:08 am

Thu Jan 03, 2019 1:18 pm

Fellside wrote:
Wed Jan 02, 2019 3:13 pm
Peter

Have a look at your first screen shot you posted.

Software Settings
Astrometry.NET
Third box down Solver URL
Remove what you have in the box.
Paste http://nova.astrometry.net

Hopefully that should get you sorted.

Graham
[/quote


Graham, thanks for your reply. I changed the URL as per your suggestion and I got immediately a pop-up window with the follow info: PLUG-In ASCOM. Astrometry Net switch not installed. I Googled this request and found a problem existing with this switch on the French website of PRISM (http://www.prism-astro.com/forum/viewto ... =24&t=1658), which translated showed me that other users had a similar problem. When I go into my Programfile(x86) >Common file>Ascom>Switch, I find there the file ASCOM.AstrometryNet.Switch.dll.

So I am not sure why the plate solve program is telling me this switch is missing..........I have the trial version of PRISM.....could there be a difference? Maybe Hamza could shed some light on this. Thanks
eurolad19
Posts: 8
Joined: Mon Nov 13, 2017 9:08 am

Thu Jan 03, 2019 1:33 pm

Hamza Touhami wrote:
Wed Jan 02, 2019 10:21 pm
Peter, if you succeed at uploading those images somewhere accessible, I'd be glad to take a look on my machine. For those interested, I will start a new series of video tutorials and I will address one simple subject in less then 2 minutes.
Hamza I would be happy to up-load this photo if you tell me where you want it. Would it make sense to up-load it here?

My other test for plate solving was with Plate solve from unknown position and used the GSC-ACT option. Despite having filled in the RA and DEC co-ordinates and the proper Focal length and pixel size of the Camera the system to over 80 minutes that it could not solve it and asked me to increase the search distance. I first started with 15 and then increased it to 30 degrees, With the same result.
Screenshot (42).png
I feel there needs to be a more in-depth explanations for the diverse possibilities of the plate solve solutions as I am really struggling to get a simple photo to solve. Thanks for your help and I hope I am not the only one with these problems.
Hamza Touhami
Site Admin
Posts: 564
Joined: Sat May 20, 2017 6:05 pm
Location: Phoenix,AZ
Contact:

Thu Jan 03, 2019 11:13 pm

We already provide an extensive number of video and written tutorials. Could there be more, sure. The allsky plate solve is designed to search for a solution all over the sky. just create a free dropbox account and upload your fits image to it, then share the link to that file with us here, we will find a solution :)

Or just email me the file directly at support@hyperion-astronomy.com
Hamza Touhami | Site Administrator
Hyperion-Astronomy.com
sponka
Posts: 12
Joined: Tue Apr 17, 2018 11:33 pm

Wed Mar 20, 2019 1:37 pm

Hi,

are there any further solutions? I have the same problem and I can't create a pointing model for my Cem120ec2 because I have to use the internal platesolving that does not work. I have all the catalogs on my hdd and set up the paths correctly.

No pointing model and no automatic focus possible because the platesolving does not work.

Kind regards Stefan
cioc.adrian.dan
Posts: 33
Joined: Mon Feb 18, 2019 1:03 pm

Mon Jul 08, 2019 10:02 am

At first platesolving was problematic for me too. After some experimenting I found that bin 2x2 or 3x3 is the way to go. I take an image binned and then right click on the image and left click use software embeded function if I recall corectly. That prompts the platesolving screen. Just press Ok (i use ucac4) and it plasesolves in 2 seconds. After that sync the telescope and thats it. It is way faster than my local copy of astrometry.net but that may be faster if i choose only the indexes that I need (i donwloaded them all :)) )
jason
Posts: 44
Joined: Thu Apr 18, 2019 3:53 am

Sat Jul 13, 2019 3:04 pm

FWIW I setup a stand alone astrometry.net instance on my laptop on windows using WSL... platesolves w/ my 183mm @ 518mm FL are now only about 10 seconds long and more robust than ever. It was a bit involved and I intend to add a tutorial for it at some point, but it improved my prism and nightly experience significantly.
rottielover
Posts: 128
Joined: Thu Aug 23, 2018 8:22 pm
Location: Missouri, USA
Contact:

Mon Aug 19, 2019 8:05 am

Jason - I would be interested in seeing that tutorial whenever you have a chance to work on it. Like Dan I've been able to get the embedded function to work and once it extracts the stars it solves nearly instantly, but there are occasions when it will fail (I think this has to do with use error on my part or something going on in the exposure, etc) and I need to fail-over.

My online solving with my astrometry.net works great, but it's a tad slow in comparison. I was able to install ANSVR, and I can even get it to solve now, albeit very slowly in comparison (usually in terms of 2-4 minutes). I think there are probably some settings I can do to ANSVR to make it faster but I'm hitting a wall when trying to find good documentation on it.
cioc.adrian.dan
Posts: 33
Joined: Mon Feb 18, 2019 1:03 pm

Wed Aug 21, 2019 12:21 pm

For the local copy of astrometry.net to solve fast you need to identify wich index files are used for solving. Go to Astrometry Log and see what index was used. Then delete all index files you don’t need. Even so it’s not as fast as prism but it always solves. Prism fails to solve if you don”t have a good focus or you don’t have enough degrees radius for search. Off course the more you increase the degrees distance for solving the longer it can take to solve. Please note that a good focus is also important and also bining. 2x2 or 3x3 also speeds up platesolving. You can easily see the effect of binning for platesolving when you use mount modeling. I had succes with model building in prism after a few tests, but the platesolving is still not as fast as platesolve 3. My sitech controller uses platesolve3 and its amazingly fast, like 60 model points in 15 minutes, but I’m getting offtopic here :). Have a nice day
rottielover
Posts: 128
Joined: Thu Aug 23, 2018 8:22 pm
Location: Missouri, USA
Contact:

Wed Aug 21, 2019 12:57 pm

Thanks Dan. My only hesitation is that I end up using 3 focal lengths with my equipment and I have to double check but I’m pretty sure I need multiple catalogs.

667mm - my scope plus focal reducer Flattener
952 mm - my scope with a Flattener
1904mm - my scope with a 2x Barlow , now I know that it’s not recommended, but I did this with the Ring Nebula (M57) and it turned out great. Normally this is my planet and moon config

I have more than a few days of clouds on my hands so I’ll take a look at which catalogs to delete.
jason
Posts: 44
Joined: Thu Apr 18, 2019 3:53 am

Tue Aug 27, 2019 12:40 am

@rottielover I'm still figuring out the best way forward for getting it working on windows. It involves installing the windows subsystem for linux, but past there there are issues with django database models - the main point of entry for the programs is a script glued to the production version of nova.astrometry.net. I know enough django to manually hack the database entries, and they seem to break frequently, this of course only started happening a couple of days after it was working just fine. I may post what I did to the tutorials section at some point but I'm going to be working on a simplified server script that just makes it easier to get going without setting the website server in its entirety.

After that I'm going to see if I can optimize astrometry.net... I want seconds at a time for platesolves.

Er anyway I don't think the key to a fast astrometry.net instance is reducing the datasets - one thing is limiting the search radius but fundamentally knowing the approximate focal length of the image would constrain most datasets out as well, though I don't know how well this is exploited. It's helpful to search all scales too, for when you don't know focal length.
Post Reply