Grainy data after stacking with calibration frames

pcalexander

New member
Hello,

I have attached my jpegs of my stacked and calibrated Ha, Oiii and Sii data. The calibration frames that I used are darks, flats, dark flats, and bias. Whenever I execute the batch preprocessing script, I still have grainy/noisy images.

The grainy noise is present my uncalibrated stacked images, so I don't think that the calibration frames are causing it. The root is possibly my filters since I just received my astrodon filters in the mail. It is possible that cleaning them will eliminate this issue, but I would like to save this data.

The Ha is 3h (90x120s), the Oiii is 1.4h (28x180s), and the Sii is 1.4h (43x120s) all shot at 76 gain, 15 offset at -10C with a ZWO ASI1600mm.

Ha: https://drive.google.com/file/d/1VPEs5cR-CSn_FVqwOi10LsXnxHb2GAfr/view?usp=sharing
Oiii: https://drive.google.com/file/d/1IPUCQYWk4t6JjRGxDhMTnyIncj9jhVmI/view?usp=sharing
Sii: https://drive.google.com/file/d/10T0_JekyN1dzv3kzUYMpPajDkuGvL1_m/view?usp=sharing

The jpeg looks the worst for the Sii.

Are there any tips for removing this noise?

Thanks,
Phillip
 
Last edited:
seems like the files are not world-readable but have been shared with someone in particular? i can't access the files.

rob
 
i wonder if the subexposures are just too short or there are too few of them. i don't know a lot about CMOS cameras but i know people usually use different gain settings for narrowband vs. RGB. did you do any research into gain/offset for narrowband and your camera?

the Ha definitely should look better; on the others the results may not be super surprising as something like 99% of the visible matter in the universe is hydrogen, and so most nebulae are much brighter in Ha than any of the other elements.

rob
 
Yes, that is true that they could have been underexposed. I definitely need more data on the Oiii and Sii.

I did research the gain/offset settings and chose to run at a lower gain for NB. I'm new to both astrophotography and narrowband imaging. I have a set of ZWO nb filters which I have used at 200 gain and 50 offset with good results. I ordered and received my new Astrodon filters and decided to try them at a much lower gain because of the fullness of the moon.

I think that I introduced too many variables to narrow down what was causing it, or if it is to be expected. My impression is that the main contribution to the artifacts/graininess is low gain with too short of subs (they could have been at least 5 minute subs for this gain).

Here is a stacked image of 15 frames at 240s at 200 gain / 50 offset with the ZWO Ha filter for comparison:
https://drive.google.com/file/d/1anoHqlKV0NKHyUUTHbJKfzxbafU6Yq5A/view?usp=sharing

And even with the data captured that is underexposed/not enough frames, here is the final image that I captured over the weekend of NGC 6820:

As you can tell, the image didn't turn out as well as it could have because I was not able to remove that noise that I was showing.

Thanks,
Phillip
 
Last edited:
are the astrodon filters narrower bandwidth than the ZWO filters?

your processed image seems pretty good despite the circumstances. maybe just keep adding more integration time. are you shooting in LP skies? even with NB filters you still need a lot of integration time from the city.

did you try MUREDenoise? the nice thing about that script is that it relies mostly on the characteristics of the sensor to model the noise. there are some user-controllable knobs but not nearly as many as, say, TGVDenoise.

rob
 
  • Like
Reactions: EAR
Yes, the Astrodon filters are 5nm, 5nm, and 3nm for Ha, Sii, and Oiii. The ZWO filters are all 7nm.

I was shooting from a borttle 4/5 area, but... under almost a full moon in a region that was way too close to the moon. The moon was near the lagoon nebula and I was shooting in Vulpecula. So, a lot of bad choices. Definitely not the filters' fault at all.

I did not try that denoise, but I will give it a try.
 
Back
Top