Author Topic: Non Linear PostProcessing  (Read 1346 times)

Offline oldwexi

  • PixInsight Guru
  • ****
  • Posts: 625
    • View Profile
    • Astronomy Pages G.W.
Re: Non Linear PostProcessing
« Reply #15 on: 2019 August 20 14:12:27 »
Let me add that the change in the Bayermatrix postion happens in my opinion
in your capturing software,  which rotetes the image by 90 degress and therefore changes in the image the bayer mnatrix from
G .. G to
. GG .

Gerald
P.S.: Just Checked the FITS Header of your image which clearly says:
Mirror in Y                   TRUE
QTURN 90 Degrees     TRUE

Setting both values from T like true to F like False should give you the bayermatrix
as mentioned by ZWO  !!!!

So dont rotate and mirror in the capturing software and everything will be fine and easy
« Last Edit: 2019 August 20 14:33:04 by oldwexi »

Offline dld

  • PixInsight Addict
  • ***
  • Posts: 128
    • View Profile
Re: Non Linear PostProcessing
« Reply #16 on: 2019 August 20 14:16:28 »
I agree with Gerald, GRBG seems to be the correct pattern. And yes the image is sharp, the camera seems fine, and I am jealous :laugh:. These discrepancies are due to how the various acquisition programs decide to arrange the pixel values into an array of numbers and then into a (FITS) file. 

PS: Take a good read at Rob's post at Cloudy Nights. I think Bernd has some nice posts somewhere in our forum explaining what's going on, but I can't find them at the moment.
« Last Edit: 2019 August 20 14:30:09 by dld »

Offline mhambuch

  • Newcomer
  • Posts: 8
    • View Profile
Re: Non Linear PostProcessing
« Reply #17 on: 2019 August 21 11:52:28 »
Hi Gerald, dld,

Thank you so much for those information, i was not aware that rotation changes the bayer matrix. I am using PRISM (like all french guys do) and i use to align the ccd with the sky in order to have the pic aligned north/east. Also happy to hear that the pic is sharp (beginners luck..) as i only started to master my equipment (esp. focus) this year. Wish you both good continuation and again thanks for your time.

Marc


Offline dld

  • PixInsight Addict
  • ***
  • Posts: 128
    • View Profile
Re: Non Linear PostProcessing
« Reply #18 on: 2019 August 21 13:27:16 »
You are welcome Marc,

Imagine an empty chessboard and yourself ready for a chess game. At the top left corner you see a white square and right next to it a black. If you rotate the chessboard by 90 degrees, the top left square is now a white. In chess there is the convention (rule) that the board is placed so that a white square is in each player's near-right corner. Unfortunately we haven't any rule for the placement of the contents of a Bayer matrix into a file, and furthermore, we have three colors instead of two!

That's why it is important for the software reading such files to know where to start reading pixel values and what color to assign at each pixel.

Any kind of luck is welcome :D even beginner's luck! I am sure you'll obtain fine images with your gear! I wish you more luck and clear skies!

dld

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4630
    • View Profile
Re: Non Linear PostProcessing
« Reply #19 on: 2019 August 21 13:29:56 »
the chessboard is a great analogy. thanks for that.

rob