Asi 1600 mc c debayer pattern

[Scott]

Robin I would just like to thank you for all your time and effort creating such an excellent product. I use 3.0 but I have been playing with your 3.1 and your additions are fantastic, especially your new histogram. Months of research on my part to understand snr and correct exposures for maximum DR is done with a push of a button. Just great. The question I have concerning my camera, Asi 1600 mc c is the debayer pattern. Sam at ZWO says that the correct pattern is grbg but the fits header is showing it as bggr. I have experimented with both and I seem to get the best color results with grbg, but with bggr the capture has less noise and cleaner looking but the color seems off or missing. With stacks of over a hundred the noise is basically gone using grbg. I'd like to know if I should just ignore the fits header and go with grbg. I do not know what pattern Sharpcap is using for it's raw capture, if it's grbg and is just posting it as bggr. I'd like to make sure that I'm using the correct pattern in my workflow as the camera recorded it because I have no idea what magic is going on in the background. Once again I thank you for a great product.
Thanks,
Scott

[turfpit]

Scott

Have a read through viewtopic.php?f=19&t=305. Kevin settled on BGGR in the end after testing all 4 patterns and processing in StarTools.

Watch out for 'flip' as that could change the required pattern.

Dave

[admin]

Hi Scott,

You're right to pick up on this but it's a bit difficult to know what to do about it as different processing programs have different ideas about how Bayer patterns apply in FITS files. For instance I can load the same FITS into PixInsight and AstroArt and in PI I need to use RGGB and in AstroArt GBRG

The cause of this is that FITS files are supposed to start at the bottom left (most images start top left). If this is followed then the bayer pattern for a FITS files (which describe the top two rows) should be swapped round in a FITS file compared to an SER file. This is what SharpCap does and it seems to be compatible with AstroArt and PIPP, but it disagrees with the way PixInsight expects the bayer pattern to be.

As you can see, this isn't something that I can fix 100% as if I swapped the pattern I wrote to the FITS file to make it what PixInsight expects then it would be wrong for AstroArt and PIPP. Maybe the best I can do is add a comment to the FITS header saying that the opposite pattern might be needed in some software?

cheers,

Robin

[Scott]

Robin I thank you fot your reply. I've been playing all day and I saw the exact same conclusion. In sharpcap the bayer offset was x1 and y1 with zwo's native driver. With the Ascom driver it was the same. In Nebulosity (I only use it now for pixstats to find exposure over background noise) I took some exposures with Zwo's native driver and had an offset of y1, x0 different than sharpcap but when I used the Ascom drive in nebulosity The offsets are y0 and x1 different pattern within the same program I came to the coclusion that it was software dependant. I use pixinsight so I'll use the grbg. Thank you kindly for your quick response and explanation. I think your response will clarify for a lot of people of what is actually going on.
Thanks again for a great program,
Scott

[Scott]

Follow up. I'd like to thank Dave and Robin for the explanation of how debeyer patterns are read, some reading bottom up others, top down. I did further testing and verified that grbg is the correct pattern in pixinsight using Sharpcap for capturing by taking daylight pics of all debeyer patterns in Pixinsight. Low and behold grbg is the correct pattern, reading opposite of Sharpcap. It all makes sense now. Thanks again.

[admin]

Hi Scott,

thanks for reporting back and confirming that you now have it working. I have made an update to the FITS file writing so that the bayer related headers now look like this

Code: Select all

    BAYERPAT= 'GBRG    '           / NOTE: Use RGGB on some software (eg PixInsight)
    COLORTYP= 'GBRG    '           / NOTE: Use RGGB on some software (eg PixInsight)

Hopefully this should help limit confusion in future (note this came from a different camera that requires RGGB for PI and GBRG for other software).

cheers,

Robin