New solve-goto feature failing to plate-solve from menu-bar

[BobHehmann]

Dr. Glover, I'm running the latest version of SC Pro, 4.0.9132, on a Windows 10 Pro laptop. ASTAP is the plate-solver supporting SC, and currently has no problem correctly performing any of the classical SC plate-solving options, such as a Scope Control's solve/refine/synch, or an all-sky solve.

A couple of days ago, I took my intended reference image (FITS) using SC to capture from an ASI071MC Pro (OSC) camera, a RedCat-51 scope, through an Antlia dual-5nm Ha/O3 filter. I'm exposing for 10 minutes, Gain 150 / Offset 20 - figures close to SC's recommended values from its exposure analysis. This are my normal production settings for bright HA nebulas, such as this target, the Pelican/NorthAm nebular region. As best I can tell, there are no problems with the FITS header's metadata. This reference exposure was captured using the immediately prior SC release, which supports Goto Image from the menu-bar.

When I submit the reference image back to current SC using the menu-bar's Goto Image, SC reports back a generic ASTAP "Unable to solve..." message, similar to what several others have reported over the last few days. But if I drop that same FITS file directly into ASTAP and ask ASTAP to solve it directly, it does so affirmatively and correctly in <0/5 seconds.

What information should I gather to help resolve this issue? Cheers, Bob

[admin]

Hi Bob,

best to send me the problem FITS file. ASTAP probably solves instantly because it sees that there are already co-ordinate headers in the FITS file and basically decides it hasn't got any work to do. SharpCap deliberately stops that info from getting to ASTAP to deal with the case where the recorded co-ordinates might not be correct. That forces Astap to do the work of plate solving, but obviously something is not quite right.

The file will be too big to upload here - the easiest thing to do is to upload it to a file sharing site (google drive, Onedrive, dropbox, etc) and share the link here to download it.

thanks,

Robin

[BobHehmann]

https://drive.google.com/file/d/1CTaIGy ... sp=sharing

I experimented a bit based on your clue - I created a version where I stripped all six Keywords that define position from a test-file's header: RA/DEC, OBJRA/DEC; OBJAlt/AZ - and by default, stand-alone ASTAP immediately rejected solving this, as its internal default was still a local spiral-search. When I set that to a 180 search, ASTAP quickly succeed solving the "stripped" file.

The google-docs link above is the original version as crated by SC (not the edited test variant).

Cheers, Bob

As an interim, "Ghost Overlay" served well...

[admin]

Hi Bob,

do you have the option 'Always send 8 bit images to the plate solving application' set in SharpCap? If I unset this then everything seems to work fine, if I set it then the plate solve fails (probably because the reduction to 8 bit throws away fine detail and means that stars are not being detected very well).

When I turn that off, I get the following key lines in the log

Code: Select all

Initializing PngFileWriter at 4944x3284x1x16bits, Bayer_RGGB, 2 bytes per PEL for       
Field of view calculated as 5.41x3.60 degrees based on FL of 250.0mm and sensor size of 23.63x15.70mm
Running external command : c:\Program Files\Astap\Astap.exe -f "C:\Users\robin\AppData\Local\Temp\e50b24ac-672d-404b-8bdd-1a45624c03d6\frame.png" -ra 20.979 -spd 133.530 -r 15.000 -z 0 -log -fov 3.60
Plate solver completed with no error code.    

Astap output in the SharpCap log

Code: Select all

    Using star database V17
    Creating grayscale x 2 binning image for solving/star alignment. 
    205 stars, 152 quads selected in the image. 136 database stars, 101 database quads required for the square search field of 3.6°. Search window at 200%
    67 of 67 quads selected matching within 0.007 tolerance.  Solution["] x:=3.978178*x+ 0.113254*y+ -9463.575238,  y:=-0.112500*x+ 3.978086*y+ -6263.556208
    Solution found: 20: 57 53.4   +43° 31 36   Solved in 0.2 sec.  Δ was 9.3'.     Used stars down to magnitude: 9.5

cheers,

Robin

[BobHehmann]

Bingo! 8-bit was checked, so I turned it off and it worked like a charm. Thanks for this feature - it is great! I tend to return to the same target 10 or more nights for an imaging run, and this makes it extremely time-efficient and easy. Cheers, Bob

PS - sorry for my delay in responding, I missed your analysis due to a brain-freeze!