solving fails in some specific sky areas, succeeds in others. repeatable

Questions, support and discussions
morrienz
Posts: 6
Joined: 31 Dec 2019, 20:20

solving fails in some specific sky areas, succeeds in others. repeatable

Post by morrienz » 31 Dec 2019, 20:36

I am trying to build a sky model for my 10Micron telescope mount using "Model Creator2" software by Astromi, and choosing its option to use Astap as its solver. I am in the southern southern hemisphere, in New Zealand. Model Images taken in the northeast of my sky solve but images in the rest of the sky all fail to solve in Astap via Model Creator. Any ideas on what might be wrong or what I can change?

I am using a ZWO ASi294MC Pro camera cooled to -15 deg C on a C11 SCT (280mm aperture) reduced to a very fast f/2 with Hyperstar, so my focal length is 560mm. It is a very fast system so doesn't need long exposures for any purpose. For this solving I tried exposures ranging from 0.75 to 3 seconds, all of which would solve in the northeast sky but not elsewhere. With Model Creator capturing and sending the images to Astap for solving I am not even sure if there are any settings/parameters in Astap that I can change that model creator won't override. I installed the G17 star catalogue for Astap.

Any help/suggestions would be welcome. It seems odd that I am getting succesful solves in part of the sky but not other. Is astap set up/tested to solve the southern sky correctly when used in the southern hemisphere. I ask that because it seems to be solving the images where I am pointing at the northern sky but perhaps not when pointing at southern hemisphere sky objects (with negative or close to negative declination).

I will send a similar message to the Astromi/Model Creator support area too, but since images solve in one sky area I suspect my problem might lie more with Astap than Model Creator.

Regards,
Chris M

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by han.k » 31 Dec 2019, 22:09

Hello Chris,

The is no difference for northern and southern hemisphere. F/2 and 560 mm is a fast system. Could you provide some images, so I can have a look.

There some things to check:

1) Are the stars round?
2) Are the stars larger then one pixel? One pixel is ignored as hot pixel.
3) What is the field of view. It could help to reduce the FOV to maybe 5 degrees. There is a setting in ASTAP for that.
4) Does a longer exposure help?
5) Is the FOV correct? If you solve an image in ASTAP, does the FOV in the statusbar change after solving. This change should be less then 5% for best performance
6) Is downsampling in ASTAP set at 0=automatic?

Please give me the link to some images. That helps to investigate.

Clear skies, Han

morrienz
Posts: 6
Joined: 31 Dec 2019, 20:20

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by morrienz » 01 Jan 2020, 00:52

Many thanks for the fast reply Hans.

I just found out that Model Creator deletes the fits images it passes to Astap each time it restarts and I restarted the program after doing the runs last night, which deleted the tif files, so i will redo it all tonight (or next clear night) and I will keep and send you some of the images that fail and some that solve. I will also send you the command strings that Model Creator passes to Astap for those images.

I will try myself to solve some of the images from next run individually in Astap as I might be able to see the problem myself. That is what the developer of model creator2 has suggested I try next, after I reported this problem to him as well.

1) Are the stars round? Will check the tifs after next run, but brighter stars will have diffraction spikes from the camera cables that run across the aperture in an SCT/Hyperstar setup since camera is sited at front of aperture. Images from this setup always solve at any sky location just fine using ZWO Asiar image capture device's solving

2) Are the stars larger then one pixel? One pixel is ignored as hot pixel. They will be larger than 1 pixel (and ZWO Asiair capture device can solve from this setup on all images at all sky locations with a 1 or 2 sec 1x1 binning exposure) but I will check after next run.

3) What is the field of view. It could help to reduce the FOV to maybe 5 degrees. There is a setting in ASTAP for that. FOV is 117.9 arcmins (just under 2 deg) x 80 arcmins (1.3 deg) . Cam sensor is 4144x2822 with 4.63 pixels., Scope is 11 inch ( 280mm) aperture with 560mm focal length. I tried both 1x1 and 2x2 binning, but got same result, images in one sky area solve, others don't.

4) Does a longer exposure help? I think I got the same result (same areas of sky failed or solved) with attempts from .75, 1, 2 and 3 second exposures time but I will try 5 seconds or more next run. It is a very fast optical system so too long and I think background sky glow will overwhelm stars.

5) Is the FOV correct? If you solve an image in ASTAP, does the FOV in the statusbar change after solving. This change should be less then 5% for best performance. Will check this on next run if I can (but I dont see the full Astap window and I don't run Astap during these runs, I just see a very small Astap solving window that opens briefly as it solves once Model Creator sends a command to Astap.)

6) Is downsampling in ASTAP set at 0=automatic? I hadn't changed any of Astap's default settings. I just installed it and the G17 star catalogue and then let model creator send commands to it. After next run i will show you the command string model creator sends to Astap. I presume the strings are correct or none of the images would solve in a run, and n one sky area they do solve.

Regards and many thanks for your help, and for developing this software for us to use.

I will gather data you have requested from next runs

Chris M

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by han.k » 01 Jan 2020, 08:42

>>5) Is the FOV correct? If you solve an image in ASTAP, does the FOV in the statusbar change after solving. This change should be less then 5% for best >>performance. Will check this on next run if I can (but I dont see the full Astap window and I don't run Astap during these runs, I just see a very small >>Astap solving window that opens briefly as it solves once Model Creator sends a command to Astap.)

Hello Chris,

You could run ASTAP as a program and load & solve the images. It has a graphical user interface. In the ASTAP viewer you can load and solve the image.

An exposure time of 1 or 2 seconds exposure should be enough for the ASI294 with F/2.

I will wait for your images.

Clear skies, Han

morrienz
Posts: 6
Joined: 31 Dec 2019, 20:20

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by morrienz » 01 Jan 2020, 11:01

Hello Han,

I have repeated the exercise tonight with exactly the same result, and same result whether I use a 0.5 second, or 1 second, or 5 second exposure, and whether or not I use 1x1 or 2x2 binning. The same sky points solve or fail to solve each time.

I should mention that the cam is a one shot colour cam with an RGGB Bayer pattern.

The files that fail to solve when passed to Astap by Model Creator also fail to solve when I load them manually into the GUI of Astap, and the files that solve in the auto process also solve when I load those fits manually in Astap.

What is the best way to send you two fits files, each is 22 MB. I can send you one that solves and one that doesn't, from the same run of model creator that tried to solve 21 images. 6 files solved, all in the northern sky, and 15 failed. I can send you the logfile from model creator which shows what it passed to Astap. The solved file I will send is the first image file for the mount model creation run, and the failed file is the second image of the run.

Regards,
Chris Morriss

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by han.k » 01 Jan 2020, 19:08

Hello Chris,

You could upload them all to the site below and give me the link. No email required.

https://uploadfiles.io

Or share them with google drive.

Maybe you star database is incomplete (290 files of *.290) but that will quickly become clear.
For a colour cam normally binning 2x is the best option.

Regards, Han

morrienz
Posts: 6
Joined: 31 Dec 2019, 20:20

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by morrienz » 01 Jan 2020, 20:04

Hello Han.

There are 290 G17 files in the Astap folder so I don't think that is the problem.

I have just PMed you a link to a shared folder in Google Drive with the various files in it, including one solved fits image from northern sky and one failed fits from southwest sky.

Thanks again for your help with this.

Chris M

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by han.k » 01 Jan 2020, 20:49

Hello Chris,

The problem is clear. The declination is stored in the FITS header with a double minus sign. That's why (far) below dec=0 it will not solve.

OBJCTDEC= '--36 40 19.6' / Nominal Declination of center of image
OBJCTRA = '23 26 33.3' / Nominal Declination of center of image


This double minus sign will result in a DEC conversion error and result will be 0, so probably below DEC=-30 degrees solving will fail.

If you load the manually the file in ASTAP you will see the red error message for DEC. If you remove one of the minus it will solve without problems.
So this is a bug in Model creator. You have to ask the the programmer of Model creator to fix it.

I will create for you a special ASTAP version to bypass this problem. Give me one hour.

Han

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by han.k » 01 Jan 2020, 21:17

Hello Chris,

A special version of ASTAP is available for 30 days at:

https://ufile.io/p4fx8vxj


That should solve your problem. Please report the the double minus sign problem to the Model Creator forum.

Clear skies, Han

morrienz
Posts: 6
Joined: 31 Dec 2019, 20:20

Re: solving fails in some specific sky areas, succeeds in others. repeatable

Post by morrienz » 01 Jan 2020, 21:29

Excellent thank you Han. That is truly wonderful support!

I will install that version and try next clear night. I will also inform the Model Creator 2 developer so he can fix that problem in the fits headers he is creating.

Many thanks again

Chris M

Post Reply