** Warning: No correlation between the master dark and target frames (channel 0)

[Rainer]

Hi,

On one night I made a session for LRGBHa filters on Simeiz22.

No I was able to calibrate the LRGB subexposures but when it comes to the Ha I get the above error.

I use the same Dark for all. Why would H-alpha from that night be any different then LRGB ?

Reading the fits header from the Ha image has the same characteristics as the LRGB images and everything is correct in there.

Any tips what could be possibly happening ?

Thanks and regards Rainer

[pfile]

if the master dark is uncalibrated, try removing the bias frame and unchecking "optimize" and see what happens to the calibrated frame.

it wouldn't be the first time that the bias frame is somehow bad and is oversubtracting from the dark.

rob

[bulrichl]

Hi Rainer,

the following suggestions not only apply to the calibration of H-alpha images, but generally. However, images captured with narrow band filters usually contain little signal. So there is an increased risk that their histogram will be clipped during the image calibration process.

I am assuming that your MasterDark was created by simply integrating the dark frames, i.e. the MasterDark is NOT pre-calibrated. In this case your settings for ImageCalibration are incorrect, please proceed with 2).

However, if your MasterDark is pre-calibrated, it might be that the MasterDark is clipped. Inspect the MasterDark
- with HistogramTransformation: check the histogram (horizontal zoom set to about 50) and
- with ImageStatistics ('Unclipped' disabled): check count (%), it should be very near 100 %.
If the MasterDark is clipped, please try the following approach:


1) Prepare a new MasterDark by simply integrating the dark frames (this is the recommended procedure, see Vicent's tutorial https://www.pixinsight.com/tutorials/master-frames/ ), i.e. the MasterDark should be NOT pre-calibrated.

2a) For using dark frame optimization, both options 'Calibrate' and 'Optimize' have to be enabled. This applies generally (i.e. also for the calibration of your LRGB frames).

2b) If you don't want to apply dark frame optimization, a MasterBias is not needed for the calibration of the light frames. In this case, disble the 'Master Bias' section. In the 'Master Dark' section, both options 'Calibrate' and 'Optimize' have to be disabled.

3) After following the above suggestions, you should inspect the resulting calibrated light frames
- with HistogramTransformation: check the histogram (horizontal zoom set to about 50) and
- with ImageStatistics ('Unclipped' disabled): check count (%), it should be very near 100 %.
The important question is: are the calibrated light frames clipped? If this still is the case, you will have to apply an output pedestal in the ImageCalibration process (in the 'Output Files' section of ImageCalibration).

Bernd

[Rainer]

Hi Rob and Bernd,

I made a second test with an uncalibrated DARK frame and one image calibrated and two did not ...

The session was made with 3203 mm focal length and each subexposure has 900 seconds of exposure time.

On the other side I have the same object taken with 979 mm focal length and there the H-alpha does calibrate trouble free with a calibrated DARK.

Will read your posts a few times so I understand them and test what you wrote.

Normally I use calibrated Master DARK and so I just do not check Calibrate in the menu.

Regards and Grüße Rainer

[niteman1946]

The warning for my setup started in 1.8.8-4 or maybe sooner.  It could be the purple text that draws more attention.

My setup has served me well for most of 9 years.  It's based on Vincent Peris' recommendations early on.

My interpretation for the Atik 383L+ mono CCD:
1.  Bias, Darks and Flats are each only integrated.
2.  Lights are calibrated with --
  a)  Master Bias (Calibrate unchecked)
  b)  Master Dark (Calibrate n Optimize checked)
  c)  Master Flat   (Calibrate checked)

This now generates the Correlation warning. 

If I un-check "Optimize" under Master Dark, the warning does not appear.  Statistics on the two resulting calibrated Lights are similar, but not identical.

1.  Should I just ignore the warning?
2.  Based on Bernd's recommendation, should I just dump the Master Bias.  Is this good practice?

Thanks Guys,
Mark

[bulrichl]

Hi Mark,

according to your description you are not following Vicent's recommendations for the preparation of the MasterFlat (your steps 1. and 2. c). In his tutorial, Vicent recommends to calibrate each individual flat frame and integrate the calibrated flat frames to the MasterFlat. If an in this way calibrated MasterFlat was prepared, it is not correct to check the option 'Calibrate' in the 'Master Flat' section during the calibration of the light frames.

Please produce a new MasterFlat by the following procedure:
- Calibrate the individual flat frames with the MasterBias, then
- integrate the calibrated flat frames to the new MasterFlat.

Use this new MasterFlat for the calibration of the light frames using your settings, except that in the 'MasterFlat' section the option 'Calibrate' is unchecked.

Please report whether the warning "Warning: No correlation between the master dark and target frames" still appears.

Bernd

[niteman1946]

Thanks Bernd,

Will do, and I'll let you know.

Mark

[niteman1946]

Bernd, a couple of things.

Please note that my standard practice for the BIAS, DARKS and FLATS is that they are only Integrated.  They are NOT calibrated and then integrated.

Consequently, I select for the Lights Calibration ----
MASTER BIAS to be "Calibrated Unchecked"
MASTER DARK to be "Calibrated Checked and Optimized Checked"
MASTER FLAT to be "Calibrated Checked"

I'm repeating myself, but just want to make this clear.

Now, having the above, I followed your recommendation on FLATS.
This time the Flats (only) were calibrated with the Master Bias.
The calibrated Flats were then integrated into a new Master Flat.
The Master Bias and Master Dark were only integrated as stated above.

So then for the Lights calibration:
MASTER BIAS to be "Calibrated Unchecked"
MASTER DARK to be "Calibrated Checked and Optimized Checked"
NEW MASTER FLAT to be "Calibrated UNCHECKED" per your instructions.

The results were the same.  The "No Correlation Warning" was the same.

Hope this is not too confusing.  I had trouble just typing it up.

Regards,
Mark

[bulrichl]

Hi Mark,

I understood your description in reply #4 correctly. The latter approach for preparing the MasterFlat is the right one, so you should keep it.

However, the result of the new light frame calibration ("No Correlation Warning") is indeed confusing. Can you please upload one bias, dark, flat and light frame each and your MasterBias, MasterDark and new MasterFlat to a filehoster and quote the link here? I would like to take a closer look at it.

Bernd

[niteman1946]

Bernd,

Will do.

 Thanks for your patience.

Mark