Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Has my QHY168 developed a fault?- advice sought.

  • Please log in to reply
5 replies to this topic

#1 pyrasanth

pyrasanth

    Gemini

  • *****
  • topic starter
  • Posts: 3,105
  • Joined: 08 Jan 2016

Posted 26 March 2020 - 07:17 AM

I moved the QHY168C camera to the Celestron C14 last night & took 260 1 minute subs through the Baader IR/UVcut filter at F7.7

 

The integrated sub looked very strange with a very odd colour cast which I was not too concerned with- colour casts are not abnormal. When I tried to clean up the image it was covered in green noise.

 

When I split the raw sub I saw that the blue channel was exceptionally noisy & broken, very weak & pushed to the far right of the red & green channels. See the histogram from the unprocessed sub.

 

Please see the attached screen shot. Could I suspect that the blue channel has developed a fault?- since this is the first image through the new setup I have nothing to compare but its very odd that the L, Red & Green channels look okay.

 

I would be grateful for your thoughts.

 

Best wishes

Mark

 

Blue Channel issue.PNG



#2 happylimpet

happylimpet

    Aurora

  • *****
  • Posts: 4,905
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 26 March 2020 - 07:43 AM

The blue channel isnt separate from the other channels. Its just specific pixels. On chips with two separate readouts (like the 1600 fr example) one will do (I think) the two green pixels in the bayer array, and the other will do the red and blue. So any problem with blue should also be present in the red.



#3 pyrasanth

pyrasanth

    Gemini

  • *****
  • topic starter
  • Posts: 3,105
  • Joined: 08 Jan 2016

Posted 26 March 2020 - 07:55 AM

The blue channel isnt separate from the other channels. Its just specific pixels. On chips with two separate readouts (like the 1600 fr example) one will do (I think) the two green pixels in the bayer array, and the other will do the red and blue. So any problem with blue should also be present in the red.

Thanks for the info- I don't know much about the colour mechanics of the camera. So- could it be that the blue needs far more integration time. I'm only testing the QHY168C on the C14 until I get the ASI6200 which I hope will be delivered in the next couple of weeks. I think OSC on the C14 at F7.7 under polluted sky is not going to be very good. I've rolled the Atik 460 EX back into service- I will live with the smaller sensor size for a few more weeks.

 

The Moravian 16200 has been sold to help fund the IMX455 camera.



#4 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,732
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 26 March 2020 - 09:04 AM

Mark,

 

   Your screen-shot seems to show that this was a color channel separation of your integrated image. That would imply that this is from a calibrated image. If so, I would not assign too much significance to the Blue Channel being brighter (farther to the right in the histogram) than the other two. That could just be telling us that the blue channel was weaker in the Flat used for calibration. That in itself is not a cause for concern.

 

   It might be interesting to see the CFA Split channels of a single uncalibrated raw frame out of the camera. I have a bit of a suspicion that you may find the Blue and one of the Green channels are fairly different when compared to the Red and other Green channel. If they all look mostly the same, then something else must be going on here. (Also try a CFA Split of a single raw Flat Frame to see how the Bayer channels look there. It could help point to what is happening here.)

 

 

John



#5 pyrasanth

pyrasanth

    Gemini

  • *****
  • topic starter
  • Posts: 3,105
  • Joined: 08 Jan 2016

Posted 26 March 2020 - 10:24 AM

Mark,

 

   Your screen-shot seems to show that this was a color channel separation of your integrated image. That would imply that this is from a calibrated image. If so, I would not assign too much significance to the Blue Channel being brighter (farther to the right in the histogram) than the other two. That could just be telling us that the blue channel was weaker in the Flat used for calibration. That in itself is not a cause for concern.

 

   It might be interesting to see the CFA Split channels of a single uncalibrated raw frame out of the camera. I have a bit of a suspicion that you may find the Blue and one of the Green channels are fairly different when compared to the Red and other Green channel. If they all look mostly the same, then something else must be going on here. (Also try a CFA Split of a single raw Flat Frame to see how the Bayer channels look there. It could help point to what is happening here.)

 

 

John

Hi John,

 

Thanks for the information you provided. The screen shots were taken from an integrated & calibrated stack of 260 1 minute subs. I debayered one of the raw subs & the histogram is below:

 

Single captured Sub.PNG

 

I then debayered the calibration master as seen below

 

Flat Frame Used in Calibration.PNG

 

However I don't know how to interpret if this is expected as the calibration channels appear to widely separated with the green & blue being very weak- is this an issue with this calibration master which may need redoing.

 

Let me know your thoughts. 


Edited by pyrasanth, 26 March 2020 - 10:25 AM.


#6 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,732
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 26 March 2020 - 11:03 AM

Mark,

 

   Those results are reasonably normal and about what I expected to see. The weak Blue channel in the Flat is what caused the higher Blue Channel in your integration. However, that is fairly normal and not a problem in itself. The Flat did just what it was supposed to do.

 

   In order to see if there may be any problems with the camera, the next step is to look at the individual Bayer pixel results from a raw un-calibrated frame. Type doing this:

  • Open a Raw Frame from the camera in PI.
  • Use the process "SplitCFA" menu selection "Process | Color Spaces | SplitCFA".
  • Uncheck the "CFA sub-folder" box in the SplitCFA process window.
  • Drag the SplitCFA process "New instance" icon (lower left) to the raw image.
    Four new image windows will open up.
     
  • Use the HistogramTransformation process to examine each of CFA0 through CFA3.

   I would expect that CFA0 and CFA3 will be similar with CFA0 being a little brighter (to the right on the histogram). CFA1 and CF2 should be nearly identical in their histograms. If they are almost exactly the same, then there are no readout problems with your camera hardware. You can use the Statistics Process to look at the levels of each CFA channels in more detail also.

 

   If you want to have another similar view of the color channels, you could also try these steps.

  • Go back to the same original raw light frame in PI.
  • Debayer it but use "SuperPixel" rather than VNG
  • Look at the histogram of this image.

   The new SuperPixel DeBayered image histogram should look much the same as that in your prior Post #5. Any difference would be surprising and could point to a processing problem of some type.

 

 

John




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics