SharpCap Forums

Hi Robin,

Wanted to try the new Sequencer feature "Goto Sky Coordinates of Image File", which will be a brilliant addition.
Alas, it did not get far. The log shows this ...

Info 18:39:12.788367 #1 The script being run was :
SEQUENCE
GOTO IMAGE C:\Astro\PixInsight\CentaurusA\frame_00007.fits
END SEQUENCE in void SharpCap.Sequencer.SequencerContext.OnStepError(ISequencerStep step, Exception exception)

Debug 18:39:12.788367 #1 Sequencer : Error : 'Plate solving failed to locate the image' while running step Goto the sky co-ordinates of C:\Astro\PixInsight\CentaurusA\frame_00007.fits in void SharpCap.Sequencer.SequencerContext.OnProgress(string e)
Info 18:39:12.792357 #1 The sequence is now finished.

The file "frame_00007.fits" loads and works OK via the Folder Camera, and does have RA/DEC (it was created by SC).

Any tips? Is there something to setup in Settings?

This feature available via the Tools menu will be a great asset too.
Cheers, Chris.

Hi,

the message means that your plate solving tool (Astap, Astrotortilla, etc) didn't find a plate solve solution for the file. The main SharpCap log will contain the output from the plate solving application (which may help work out what went wrong) as well as the command line parameters that SharpCap used to run the plate solver.

Please feel free to post the log here and I will see what can be worked out. I will also update that error message to make it a bit more informative.

cheers,

Robin

Hi Robin,

I'm using ASTAP, which normally works from SC 4.0. Now with 4.0.9094, I tried the new goto image tool, without "Refine position, plate-solve and sync". It returns an error trying to plate-solve. The log is attached (I switched on "General" for extra logging). I am using the Telescope simulator for mount.

It first logs:
'Image loaded, starting plate solve centered on RA=12:03:03,Dec=-18:59:33, radius of 15 degrees.'
(This would have worked OK, one expects.)

But then later:
Starting blind plate solve
Field of view calculated as 162.52x154.55 degrees based on FL of 1475.0mm and sensor size of 19186.72x13065.86mm
(Now that's what I call a "full frame" sensor! That will go with my 203 m aperture telescope!)
Running external command : c:\Program Files\Astap\Astap.exe -f "C:\Users\rya113\AppData\Local\Temp\6d665da2-8e13-46d0-8b94-3f52b7e0c96a\frame.fits" -ra 12.051 -spd 71.007 -r 15.000 -z 0 -log -fov 154.55

I then checked that ASTAP would work direct and get this log, not far away in RA/DEC ...

13:35:38 De-mosaic bayer pattern used RGGB
13:35:38 Adjusting colour levels as set in tab "stack method"
13:35:41 Applying colour-smoothing filter image as set in tab "stack method". Factors are set in tab "pixel math 1"
13:37:31 Using star database H18
13:37:31 Creating grayscale x 2 binning image for solving/star alignment.
13:37:31 413 stars, 313 quads selected in the image. 281 database stars, 213 database quads required for the square search field of 0.5°. Search window at 100% based on the number of quads. Step size at 100% of image height
13:37:31 75 of 76 quads selected matching within 0.007 tolerance. Solution["] x:=-0.003321*x+ 0.647122*y+ -905.896589, y:=0.647160*x+ 0.003341*y+ -1345.501741
13:37:31 Solution found: 12: 01 54.5 -18° 51 14 Solved in 0.4 sec. Δ was 18.2'. Mount Δα=16.2', Δδ=-8.3'. Used stars down to magnitude: 18.0

This will be really cool, when it works.
Cheers, Chris.

... actually tried it with and without the "Refine position, plate-solve and sync" option.
I think the log posted above is actually for the WITH case.
Cheers, Chris.

Hi Chris,

that's weird - the sensor size comes from resolution of the image and the pixel size, so I can only imagine that the pixel size headers in your FITS file might be a bit weird - I would expect them to be in microns, so for instance 3.75 or 4.8.

If all looks OK then please share one of the FITS files you are using in the folder monitor camera with me and I will investigate.

cheers,

Robin

Hi Robin,

One of the FITS files (frame_00071.fits) that fails is here: https://cloudstor.aarnet.edu.au/plus/s/F9ckVPXkz11EQLg

It gives a similar error and log (see attached).

Cheers, Chris.

Hi,

Ok, got it - the sensor size is mistakenly calculated in microns, not mm - if you have a focal length set then this combines with the focal length and "bad stuff" happens to the plate solving. I have tweaked the code to calculate the sensor size correctly *and* to use the focal length header from the FITS file rather than the focal length from the ASCOM mount or from the sharpcap settings - should fix things hopefully.

cheers,

Robin

Cool. I'll test it out after Monday's update.
Cheers, Chris.

Cool, why wait for Monday, now have SC 4.0.9109 to test ...

Tested the goto image tool, first using a FITS file and then a PNG file of the same object.
File is here: https://cloudstor.aarnet.edu.au/plus/s/0NhvtjXxgB7R97K

1. FITS file
The image of Centaurus A (JNOW: RA=13:26:49,Dec=-43:07:23) has FITS headers written by SC for RA=201.707111, DEC=-43.1231226, OBJCTRA='13 26 49.0', OBJCTDEC='-43 07 23.0' in JNOW (Equinox=2022.4617486), FOCALLEN=1488, XPIXSZ=YPIXSZ=4.63. The log shows it uses a starting centre for ASTAP of RA=13:26:49,Dec=-43:07:23, radius 15 degrees. This seems OK. But the log shows it using "Field of view calculated as 162.52x154.55 degrees based on FL of 1475.0mm and sensor size of 19186.72x13065.86mm". Hence, it still seems to not recognize the pixels size as microns.

2. PNG file
The PNG is made from the above FITS of Centaurus A. Interestingly, the log says it uses a hint centered on RA=13:07:06,Dec=-37:41:53, radius of 180 degrees (i.e. blind solve). Not sure where this coordinate comes from. The Simulated mount is currently pointing at Centaurus A, so not from there.
This does work, solution: RA=13: 25 31.0, DEC= -43° 00 43 (J2000).

Hence, it seems that although the FITS file is not being passed to ASTAP, just a PNG copy of it, the passed FOV is still way off due to the sensor calculation.

Cheers, Chris.

Hi Chris,

yes, you are right, I am sending a PNG copy of the FITS file to the plate solver - this solves the problem of the plate solver peeking at the existing headers and also the fact that some of the code around the plate solving function in SharpCap has been written on the assumption of PNG files. While I could change that, there are more things that could get accidentally broken that way.

On the plus side, I found the error with the sensor size yesterday and fixed it. I also made sure that the plate solve from FITS uses the FOCALLEN header from the FITS file rather than the currently set focal length (in case the two are different). Should all work on Monday's version I think.

cheers,

Robin