Odd problems only on Ha images.

[Mark de Regt]

In my last two images, I've had two odd issues crop up:

1. My calibrated Ha images are noisy, when usually they're not.  The uncalibrated images are not noisy.  No other images have this problem.  I use the same darks and biases on the Ha that I use on the other images, and the Ha flat is not at all noisy.

2. After running all images through Star Align, all Ha images (and only Ha images) have a cross-hatched pattern in them.  Each has a different one, but all have them.  I have attached a sample.  This pattern is not there in the calibrated Ha images, or the Ha images after running the calibrated images through cosmetic correction, but they are there after star alignment.

Interestingly, the cross-hatching largely disappears with a Winsorized sigma clip integration.

Any ideas?

Thanks.

Mark

[bulrichl]

Hi Mark,

probably the calibrated Ha images are clipped. Also see the discussion here: https://pixinsight.com/forum/index.php?topic=10827 or https://pixinsight.com/forum/index.php?topic=12295 .

You can easily check whether this is the case: inspect some calibrated Ha images in HistogramTransformation with a horizontal zoom of about 50, and in Statistics with the option 'Unclipped' unchecked. In the histogram, the peak must not be cut at the left, and in Statistics the count (%) value shall be very near 100 % ([100 % - count (%)] is the fraction of clipped pixels). I suggest you to take a look at the calibrated non-Ha images as well and compare.

One possible reason for clipping in the calibrated Ha images might be a wrong preparation of the MasterDark. Neither the individual dark frames nor the integrated dark frames should be pre-calibrated with a MasterBias). A calibration of the MasterDark with a MasterBias is only necessary when dark frame optimization is to be used. In this case the better approach is, to calibrate the MasterDark during the image calibration of the light frames, and not before.

However, the effect can appear as well with a correctly prepared MasterDark. Frames taken with narrow band filters often contain weak signal (due to low exposure time), and the noise can cause clipping in the calibrated frames. In this case, the application of an output pedestal (during light frame calibration, in the 'Output Files' section of ImageCalibration) will put things right.

Bernd

[Mark de Regt]

Wow, that worked!

Since I had used the same darks in all the images, and calibrated all images the same way, I thought it unlikely that they were to blame.  So I went back to the raw Ha files and added an 80-count output pedestal, and all is well.

For the record, these were all 30-minute individual exposures.

Thanks so much!

Mark

[bulrichl]

I am glad that it worked for you and I appreciate your feedback. That way, other users who encounter a similar problem have the possibility to find a solution by searching this forum, so thank you, Mark.

Bernd