Bayer Pattern Confusion

Nick Kohrn

I am using sample data to practice with PixInsight after completing the FastTrack course, and I am having difficulty when attempting to determine the Bayer pattern of the sensor used. The sample data that I am using to process IC 5146 (The Cocoon Nebula) can be found here. While looking at the information provided about the equipment used, located here, I read the following description:

"This image was captured from my Bortle Scale Class 6 backyard using a Canon EOS 60Da DSLR camera. A SkyTech CLS-CCD filter was used to help isolate this nebula from a light-polluted sky. The image was captured through a William Optics Zenithstar 73 APO refractor in June 2019."

When looking at the FITS header, I see camera listed as 'Canon EOS 60D'. So, I searched the Internet to find the Bayer pattern for the camera, and I found many sites state the Bayer pattern as 'RGGB', which seems to be typical for Canon cameras. After applying the DeBayer pattern to my sample light frame, the nebula is displayed in green coloration. However, the nebula appears in red coloration when using the automatic Bayer pattern detection. To determine what pattern is being detected by the automatic setting, I began testing and concluded that the only pattern setting that produces the red coloration is 'GBRG'. 

Is there something that I may be doing to incorrectly determine the Bayer pattern? The information that I found on the Internet does not seem to align with the results of actually applying a DeBayer function on the data.

Auto

GBRG

RGGB

BlockHead

So I looked at your data..and I can't explain why the pattern isn't RGGB... but it definitely is not.

The fact that Auto found the correct pattern I assume is due to the LibRaw thing. Discovering it on your own was good work! You can also look at the Console when you have Auto. checked..and it will tell you which pattern it used (whcih agrees with you).

So this is a good lesson to do the discovering on your own. Don't rely on something someone told you... even the manufacturer. Always confirm!

-the Blockhead

Jeff Kieft

I had a similar situation - I have a Canon 5Ti/700D that I am using to get started in imaging and based on internet searches, it seemed like RGGB was the standard Canon pattern. When I ran WBPP with Auto detect for the Bayer pattern on a bunch of images of M33, it indeed came up with RGGB and I thought all was good...but my final calibrated and integrated image was very green. 

I went ahead and processed it (DBE, NoiseXterminator, SPCC, etc.) and managed to get a reasonably decent image (for a newbie). 

But then I watched Adam's videos and decided to back up and do a better job of "sanity checking" the images before calibrating and integrating. I played with the bayer pattern and GBRG seemed better, maybe.... When I ran auto detect with the Debayer Process, it came up as GBRG so I am thinking that is the right choice,

So, interestingly, WBPP chooses RGGB. but the Debayer Process chooses GBRG. I am going to reprocess with the latter.

Lesson learned, thanks Adam for your instruction!

-Jeff

Update: I restarted PixInsight and tried again, and this time both came up with RGGB...so I am a bit confused!

BlockHead

Jeff,

All WBPP is doing is reading the FITs header. Please examine the FITs header and see what is written there. (It should say BayerPat: RGGB or something like this)

-the Blockhead