NGC7023 Iris Nebula

[Budgie1]

We had a lovely clear evening last night and I managed a good 5 hours imaging.

I managed to get a 10 3 minute subs of the Iris Nebula the evening before, the first time I'd had a go at this target, and liked the results. So spent last night on the target, the longest I've ever put into one target to date. 

This is 80 x 3 minute subs, 20x flats, 20x dark flats & 20 x bias.

Kit: SW EQ5 Pro, SW Evostar 100ED Pro with 0.85 reducer/flattener, modded Canon EOS 1300D with PHD2 and APT. I also used dithering.

Image is stacked in DSS and processed in PI. I reduced the brightness of the core to bring out some of the detail in there, which sort of worked.

I'm happy with the detail but the guiding wasn't the best last night and the stars are a bit bloated, so I've now tightened up the Dec axis on the mount to try & make it better.

[Budgie1]

Here's another version which includes some 5 minute subs I took at the same time and is processed in Photoshop. I think it shows the core much better.

I'm interested if anyone can tell me what the horizontal banding is caused by, would a larger dither remove this?

So this one is 80x 3 min Subs, 14x 5 min Subs, 20 flats, 20 dark flats & 20 bias. A total of 5 hours & 10 minutes ingestion.

[AstroNtinos]

This is not like a popular target but it's stunning. I love it. Great job.

Edited December 8, 2020 by AstroNtinos

[wimvb]

Very nice image. The banding is typical for many dslr cameras, but especially Canon. That's why PI has CBR, Canon Banding Reduction.

I think that if you use SCNR Green on the first image, this will lift the blue colours.

[Budgie1]

24 minutes ago, wimvb said:

Very nice image. The banding is typical for many dslr cameras, but especially Canon. That's why PI has CBR, Canon Banding Reduction.

I think that if you use SCNR Green on the first image, this will lift the blue colours.

Thanks Wim, the CBR did the trick. I ran quick process on it and using 0.70 on CBR script removed the banding.

I'm only about 3 months into PI and I hadn't found CBR in the script menu.

I also tried SCNR on the green and did work. When I get an evening I'll re-process the image again and see what I can get out it.

[Budgie1]

I did a complete reprocess of the TIFF using PI today. This included removing the banding with CBR and SCNR for the green.

There is still a bit of noise in there but it's better than the first attempt and this is how it turned out.

[tooth_dr]

Very nice, the banding has been removed very successfully.

[wimvb]

4 hours ago, Budgie1 said:

I did a complete reprocess of the TIFF using PI today. This included removing the banding with CBR and SCNR for the green.

There is still a bit of noise in there but it's better than the first attempt and this is how it turned out.

Excellent.

If you want to improve this further, you can reduce the colour mottle (chrominance noise) with multiscale median transform. Have a look here

http://wimvberlo.blogspot.com/search/label/noise reduction

[Budgie1]

Thanks @wimvb.

I checked your blog and ran through MMT.

I used TGVDenoise on the one above but in RGB/K mode and then used Multiscale Linear Transformation on the luminance and chrominance layers. I take it TGVDenoise should be in CIE L*a*b* mode and on lightness? Unfortunately the links on your blog page for TGVDenoise don't work now.

After going through the processing I ended up removing the TGVDenoise (as it did some strange things to the stars) and just ran MMT which reduced the colour noise. I then ran MLT on the luminance to reduce the pixalation of the noise. I've learned a lot with your assistance and with a few other tweaks, here's the result:

 

[wimvb]

I also don't use tgvdenoise any more, nowadays. This is partly because I use longer integration times from a much darker site, and partly because it introduces very subtle artefacts in my images. But the main reason I posted that link, was to show you mmt on colour noise. I tend to look at an image in terms of lightness and colour, luminance and chrominance, and I use the appropriate colour space for noise reduction. The small scale "salt and peppar" noise is usually lightness noise. I use MLT mostly, to reduce that. Larger scale colour mottle doesn't show up in the L channel. This I treat with MMT. Many people just blur the RGB image to get rid of colour noise, but I don't like that.