Possible issues with creating FITS files when using Flip Horizontal

[TDunnavant]

Two different versions of SharpCap 4.0 64-bit, on two different machines:

Machine 1 (which appears to operate correctly - this is a laptop):
SharpCap 4.0.9132.0,
OS = Windows 11 Home
Camera = ZWO ASI533MC Pro

Machine 2 (where the problems occur - this is a miniPC):
SharpCap 4.0.9208.0, then upgraded to 4.0.9224.0
OS = Windows 11 Pro
Camera = ZWO ASI533MC Pro

In all cases, attempted to take darks using Horizontal Flip and with Debayer Preview set to "On".

In all cases, log file states that SharpCap is "Initializing FitsFileWriter at 3008x3008x1x16bits, Bayer_GRBG, 2 bytes per PEL..." This is what I would expect for a horizontally flipped Fits file created using this camera.

For Machine 1, I opened the resulting Master Darks in ASIFitsViewer and saw an as-expected Fits file (at regular magnification, an appearance of a noisy view).

For Machine 2, I opened the resulting Master Darks in ASIFitsViewer and saw a checkerboard pattern.

For all sets of darks, I opened them using two instances ASIFitsViewer, with one instance running on Machine 1 and one instance running on Machine 2. The appearance of each Fits file was independent of the instance of ASIFitsViewer that was used to view it.

I attach the log file for the problematic machine (Machine 2), running SharpCap Pro 4.0.0224.0

[admin]

Hi,

can you check the FITS headers of the two created FITS files, in particular the headers 'COLORTYP', 'BAYERPAT', BAYOFFX, BAYOFFY, XBAYROFF, YBAYROFF

These headers are how SharpCap tries to communicate to FITS reading software the information about the pattern of the sensor. Unfortunately different reading software takes the information from different headers, which is why SharpCap has to write so many different options.

If all those headers are the same on both PCs then SharpCap is doing everything that it is supposed to. It's possible that ASIFitsView is one of those pieces of software that ignores these FITS headers entirely and has a setting somewhere for whether to debayer or not - if so it may be worth you checking that the setting is set the same way on both machines.

cheers,

Robin

[TDunnavant]

Will do, Robin, once I figure out how to examine the header info on Fits files.

Astrophotography is a never-ending learning process

[TDunnavant]

Hi, Robin,

Here are some extracts from the Fits header files for two horizontally flipped Fits files. I omit the meta-data for image time, camera, ccd-temp, etc.

As best as I can tell, the meaningful attributes are identical.

For the first file, taken using SharpCap 4.0.8949.0, ASIFitsViewer presents a "reasonable" Fits dark file. For the second file, obtained using a near-current version of SharpCap (4.0.9208.0), ASIFitsViewer presents a file with a checkerboard pattern.

Later today, or maybe this evening, I'll down-version SharpCap on my miniPC and repeat my experiments. This will provide additional evidence as to whether what I'm seeing is due to changes in the more current versions of SharpCap or is due to changes in the machine/OS on which SharpCap is running.

For a flipped file taken with an older version of SharpCap - in this case 4.0.8949.0 - I have the following:
- BAYERPAT = 'GRBG '
- BAYOFFX = 1
- BAYOFFY = 0
- BITPIX = 16
- BLKLEVEL = 5
- BSCALE = 1
- BZERO = 32768
...
- COLORTYP = 'GRBG '
...
- EXTEND = T
- ROWORDER = 'TOP DOWN"
- SIMPLE = T
- XBAYROFF = 1
- YBAYROFF = 0
- XBINNING = YBINNING = 1

This flipped file shows fine in ASIFitsViewers. I see a reasonable FITS file.

For a flipped file taken with a more current version of SharpCap -- in this case 4.0.9208.0 - I have the following:
- BAYERPAT = 'GRBG '
- BAYOFFX = 1
- BAYOFFY = 0
- BITPIX = 16
- BLKLEVEL = 5
- BSCALE = 1
- BZERO = 32768
...
- COLORTYP = 'GRBG '
...
- EXTEND = T
- ROWORDER = 'TOP DOWN"
- SIMPLE = T
- XBAYROFF = 1
- YBAYROFF = 0
- XBINNING = YBINNING = 1

[TDunnavant]

Hi, Robin,

I seem to have reported a red herring.

I down-revved SharpCap on my miniPC to the same version I have on my laptop. Upon creating a new Fits-based dark, I'm seeing the same "checkerboard" pattern as I saw with the more recent product versions, but which I didn't see while using this older version of SharpCap on my laptop.

So...

The problem appears to lie elsewhere in my tool chain!