11 Hours of data, heaps of noise?

Processing Deep Sky NGC5367 Nebula
15 replies•677 views
Lachlan Wilson avatar
Lachlan Wilson avatar
Hi all,


I'm sure this all comes down to light pollution but I thought it would be worth asking. I have 10 hours 50mins of data on NGC5367 which is a reflection nebula alongside a Cometary Globule. These were taken over 2 nights, and have 130x5mins exposures. I was also imaging at around Bortle 4/5.

When I put them into Pixinsights WPBB, I selected to drizzle (only 1x) because I have heard that it creates a better SNR. It all went well and produced images. When I was processing the drizzled image, I was getting very little of the nebula, and a lot of the noise. I was hoping that 10 nearly 11 hours would at least begin to drown out a lot of the noise that I often experience.

Attached is the best image that I could create out of the drizzled data, but I'm really not happy with it considering all the images on Astrobin about this nebula.

When I went back and got the linear images of the drizzled and undrizzled images, it sort of appeared that the undrizzeled image had less noise in it that the other. Why is this the case?
I would also love it if someone could have a quick glance at the images and have a little process of them as I just can't get anything that I am happy with.

Here is a link to a Wetransfer for you to download both the drizzled and undrizzled images.

Many thanks in advance


Lachlan


Oh and gear I used
Sharpstar 90mm triplet
ZWO ASI533MC Pro
130x5mins

Apologies for the image quality
Stefan Pfleger avatar
Stefan Pfleger avatar
Hi Lachlan, this is a darknebula…
I usually go for 20hrs + on these since they are very faint and need tons of integration time to resolve well. You being a B4/5 doesn‘t help in that matter. Whenever you have a lot of noise, always try to gather more data, preferably from best conditions possible, i.e. no moon.

As to drizzle making it more noisy, that is an area I don‘t know that well yet, but if the data wasn‘t dithered, or poorly dithered, drizzle will not work well.
D. Jung avatar
D. Jung avatar
You have some really difficult to remove gradients in your image. You could try to create a better flat.


 If you can remove them, you can stretch the image much more. I did a quick edit with a single DBE. the issue is that the gradients are also going through the nebula causing weird colour striations. other than that, the data is good enough for a decent edit. only used PI for the edit with Russells Xterminator plugins.
Rafał Szwejkowski avatar
Rafał Szwejkowski avatar
Drizzle will add resolution at 2x and recover some detail from the Bayer matrix at 1x but at the expense of noise.
Stefan Pfleger avatar
Stefan Pfleger avatar
Rafał Szwejkowski:
Drizzle will add resolution at 2x and recover some detail from the Bayer matrix at 1x but at the expense of noise.

That is true, but the OP posted about using 1x drizzle, its a bit different.
andrea tasselli avatar
andrea tasselli avatar
Drizzle always reduces the SNR of the final image but you get better resolution of OSC data which means you need to have a very good SNR to extract most benefits from it.
Rafał Szwejkowski avatar
Rafał Szwejkowski avatar
Stefan Pfleger:
Rafał Szwejkowski:
Drizzle will add resolution at 2x and recover some detail from the Bayer matrix at 1x but at the expense of noise.

That is true, but the OP posted about using 1x drizzle, its a bit different.

You're saying 1x increases SNR?
Stefan Pfleger avatar
Stefan Pfleger avatar
Rafał Szwejkowski:
Stefan Pfleger:
Rafał Szwejkowski:
Drizzle will add resolution at 2x and recover some detail from the Bayer matrix at 1x but at the expense of noise.

That is true, but the OP posted about using 1x drizzle, its a bit different.

You're saying 1x increases SNR?

No I never said that. I just wanted to refer to the 1x part and that it doesn't really add resolution. I think its cfa drizzle? Makes the colors a a bit better with the bayer Matrix or something? Idk i have absolutely no clue to be quite honest, mainly just wanted to point out that it was 1x and not 2x.
Lachlan Wilson avatar
Lachlan Wilson avatar
D. Jung:
You have some really difficult to remove gradients in your image. You could try to create a better flat.


 If you can remove them, you can stretch the image much more. I did a quick edit with a single DBE. the issue is that the gradients are also going through the nebula causing weird colour striations. other than that, the data is good enough for a decent edit. only used PI for the edit with Russells Xterminator plugins.

I was wondering where the gradient came from. I might process the second night of data and see if that has the gradient as well as I think it only stemmed from the first night.

I thought/knew that I would need more data and darker skies. I have a trip planned to B1 in a couple weekends time so hopefully I can get good data then (its also new moon then yay).

Thanks for the effort with a process. It looks great. I don't have Noise Xterminator so I am only using the EZ Processing suite. Haven't had a hard enough target like this yet to warrant it. Maybe I will get it soon.
Lachlan Wilson avatar
Lachlan Wilson avatar
Thanks everyone for your input. I must have read wrongly that it added SNR. Whenever I have over about 20 subs, I just tick the drizzle 1x button in Pixinsight which I have heard does wonders, but never actually worked out if it does. I might do a compare and contrast on a different image that actually worked.
Oskari Nikkinen avatar
Oskari Nikkinen avatar
That gradient reminds me of a problem i had with a previous coma corrector for my newtonian (the TS Maxfield 0.95x one) where some kind of internal reflection appeared in a rainbow fashion if a bright star was just the right distance off center outside the field of view. Looks like there are a number of quite bright stars nearby here too, and it might be related to something internal rather than a sky gradient.If so it will be difficult to get rid of.

I created a gif of an RGB and separate R-G-B stretched images. To me it looks like Nu Centauri a couple of degrees off might be to blame if this is an internal reflection of some lens surfaces within your scope.
Lachlan Wilson avatar
Lachlan Wilson avatar
Thanks Oskari. Cool GIF! I will have a look at stellarium and work it out. Next time I'm imaging, I might rotate the camera and see if it helps it in any way. Any tips for how how get rid of it any other way?
Joe Linington avatar
Joe Linington avatar
There is a posted bug in the current release of WBPP that makes a Bayer/CFA/1x drizzle not process properly. Normally this would be the best way to get good data from an OSC camera if all the preconditions are met but until this is fixed, you have to use a work around process or not select 1x drizzle. Or use APP to do a bayer drizzle.
Oskari Nikkinen avatar
Oskari Nikkinen avatar
Lachlan Wilson:
Thanks Oskari. Cool GIF! I will have a look at stellarium and work it out. Next time I'm imaging, I might rotate the camera and see if it helps it in any way. Any tips for how how get rid of it any other way?

Bruteforce gradient removal with Graxpert/Siril background extractor gets mostly rid of it. Lots of samplers and make sure none of them are on stars, but place lots of samplers on the rainbow itself. Fiddle with the smoothing value and it mostly goes away from what i tested.

Below a quick attempt, you see its still there but this level of stretch is way beyond what you will do in processing so i would probably call it good enough. You got very lucky here that the effect is not on the object of interest and is sort of easy to remove. You could do the gradient removal on a starless image too which will probably be easier since this is a rich starfield. For final touches you could try and selectively desaturate the rainbow areas with masks to completely hide it.
Chris White- Overcast Observatory avatar
Chris White- Overcast Observatory avatar
Lachlan Wilson:
Thanks Oskari. Cool GIF! I will have a look at stellarium and work it out. Next time I'm imaging, I might rotate the camera and see if it helps it in any way. Any tips for how how get rid of it any other way?



There is a possibility that something like this is caused by a shiny surface of an adapter in your image train.  Take a look at this image:  https://astrob.in/f3w5so/0/

I had an M68 to M54 adapter that was shiny on the inside.  I flocked it and it resolved the ring issue.
Baronvonsmoogle avatar
Baronvonsmoogle avatar
Wouldn't 60 second subs be a better option in a Bortle 4/5 for helping to control gradients/light pollution?
Related discussions
Drizzle/Binning/Resampling
Fellow Astrophotographers, i am having a day off and thanks to the bad weather no new data to process, so I decided to make a comparison of (the same) differently sampled luminance data of my latest project. what do we look at? https://www.astrobin.c...
Directly addresses drizzle integration and resampling techniques discussed in source.
Dec 22, 2024
Drizzle integration issue in PixInsight after two sets of images with different resolution have been previously correctly integrated
Hi everyone, today I’m presenting an issue I’ve encountered while trying to integrate two sets of images with different resolutions and original sampling in PixInsight: one taken with an ASI533MC PRO + 115/800 telescope and the other with a Canon 90D...
Covers drizzle integration issues with multiple image sets in PixInsight.
Dec 6, 2024
Help Me Understand The End Game Benefit of Drizzle
For almost a year now, maybe longer, I have been dutifully dithering my data at least every 2 subs and by 5 pixels. I have done this because…well becuase you're supposed to according to everyone. I know there are benefits beyond drizzling a...
Explains drizzle benefits and dithering strategy relevant to author's data collection.
Mar 8, 2025