Vertical Pattern Noise

[mike@mulcahys.us]

Hello all,

I am in the process of setting up a new wide field imaging system.  Specifically I am using a ZWO ASI183MC Pro with a Rokinon 85 mm lens at f/4.0.  It is a very sensitive OSC camera with 2.4 micron pixels.  My specific problem is that faint vertical pattern noise appears after image integration.  I am attaching two image links, one subframe after calibration, debayer, and registration and one image of a stack of 60 subframes.

https://www.dropbox.com/s/emea5s40tboqajj/ngc2174_1x1_120s.05109_c_d_r.jpg?dl=0


https://www.dropbox.com/s/kqvamy54p95ftu7/integration2.jpg?dl=0


The PI process was:

calibrate with bias subtracted dark frame only
debayer with RGGB Bilinear
Register with Star Alignment default settings
Integrate with default settings and sigma rejection at +- 3

This is my first effort at OSC processing in PI so I am not sure what I should be seeing.  The sub seems to show vestiges of the bayer matrix.  Is that normal?  The faint vertical banding seems to be made up of faint matrix leftovers for lack of a better description.

I have spent several days trying to figure this out with online research and trying various settings to no avail.

Any help would be appreciated,

Mike Mulcahy

[bulrichl]

> calibrate with bias subtracted dark frame only

Maybe I misunderstood you, but if you only subtract a bias-subtracted dark frame (or a bias-subtracted MasterDark) from the lights, the calibration result will be not correct (see: https://pixinsight.com/forum/index.php?topic=11968 ).

A MasterDark shall be made by integration of NOT bias-subtracted darks. The MasterDark shall NOT be bias-subtracted as well. Then there are two posibilities in the calibration of the light frames:

1) Without Dark frame optimization:
Be shure to uncheck the 'Master Bias' section. In the 'Master Dark' section select the MasterDark, uncheck both 'Calibrate' and 'Optimize',

or

2) With Dark frame optimization:
Be shure to check the 'Master Bias' section and select the MasterBias. In the 'Master Dark' section select the MasterDark, check both 'Calibrate' and 'Optimize'.

If then the problem persists, try a different debayer algorithm (e.g. AHD).

Bernd

[mike@mulcahys.us]

Hi Bernd,

Thank you for your reply.

The master dark was not bias subtracted, just a stack of dark frames.  I reprocessed the images with no calibration and the same vertical banding appears.

I also tried all the debayer algorithms offered in PI.  VNG and Bilinear both show the pattern noise with Bilinear being slightly better.  The super pixel produced no obvious pattern noise but has more color noise and some loss of resolution.  I don't know what AHD is.

The vertical banding noise appears after image integration.  The calibrated, debayered and registered images have noise in the background but not banding.  The slight differences in the intensity of background pixels are not smoothing out after integration of 60 subframes which seems to lead to the banding.

I am not sure where to go from here.  Perhaps tuning the pixel rejection in the integration process.  Other ideas would be appreciated.

Mike

[bulrichl]

Hi Mike,

I don't know what AHD is.

Yes, my fault. AHD is a debayer algorithm that is provided in FormatExplorer/DSLR_RAW or FormatExplorer/RAW, but not in the Debayer module.

Sorry, then I have no other ideas.

Bernd