FITS & Sharpcap issue

[Cosmic Geoff]

While doing some EVAA imaging of faint planetary nebulae I became aware of a problem with live stacked images stored as FITS files.  Kit: 102mm f5 achro, ASI224MC camera, EQ5 Synscan mount, Sharpcap4. 

I noticed that the colours in some of the de-bayered FITS images, converted to TIFF in the de-bayering using PIPP, looked off.  I nearly always take a companion single shot saved as a .png image. I will try to give examples below.  You will note that the de-bayered .png image has credible colours for various stars and for the nebula in the centre, which looks blue-green.  The live-stacked FITS image, hovever, debayered and converted to .tiff in PIPP, exported as a jpg in GIMP, has green stars! -which is plainly wrong.  They also look green in the PIPP preview, BTW. This is the most extreme example I have found so far - mostly the FITS colours just look less saturated or a bit off.

(Note that I am not concerned about those blue or green fringed stars as such - they appear because I have not invested in an ED or APO scope.)

So what's wrong? Clearly not the camera. After some digging, I found that Sharpcap 4, unlike the previous version, has settings in the Image tab where one can set the indication of the Bayer matrix to be correct for PIPP (etc) or Astap (etc) and unlike the previous version it defaults to the latter. 

OK, fine, so I just have to change that setting to PIPP (etc), reprocess the previous images (since I installed Sharpcap4) and the problem is solved? Actually, no...☹️.  I tried loading the NGC7026 FITS into ASTAP and found I could have either green stars (which it seemed to default to) or a perfectly monochrome image. Nothing else. That leaves two other programs (mentioned in the Sharpcap settings.  Pixinsight is a paid-for program, and My Windows 10 wanted nothing to do with opening Siril.

At which point I gave up.  Any ideas?

 

Stack_16bits_10frames_100s.tif

[Cosmic Geoff]

I had a further look into this and am more confused than ever.  I was wrong about that setting in Sharpcap 4 :

Debayering FITS with PIPP Not Working in Version 4 - SharpCap Forums

It looks like I was not using 32-bit files so this is not relevant.

I was also wrong about the camera not being relevant - I tried some shots with my ASI120MC camera of the same targets as with the ASI224MC. With the ASI120MC the .png single shot and the .fits stack come out having the same colour rendering (and the bright stars don't have green haloes).

With the ASI224MC the results are as above.

I also noted that when processing a FITS file in PIPP, it does not matter what debayering options are selected, the result is exactly the same. Maybe the stacked FITS file is already debayered?

Also noted that when processing a .png file from either camera, it is possible to recreate the 'green halo' effect by selecting a 'wrong' debayering option.

Still baffled.

[happy-kat]

Does feel like the bayer matrix in sharpcap is off. My ASI462MC is ok in sharpcap 4 and I collected raw files to stack.

[Cosmic Geoff]

The FITS stack is now working properly, and I think I have figured out the answer to this. If you set 'Debayer Preview' to ON, all is well, similarly if you set it to RGGB (with an ASI224MC).  But if you set the 'Debayer Preview' to an incorrect setting like GRBG, it looks wrong on the preview and this wrong colour balance persists in the saved stacked FITS image.