Whats this stuff?

[Gavin1234]

I think this might be related to an ongoing problem Ive asked about before.

Whenever I image something with a lot of background (Helix and NGC253) I seem to get similar problems. No such problems on any other targets e.g. Horsehead, Orion, Lagoon etc. etc.


Anyone able to tell me what this background stuff is? Conditions weren't great last night but they weren't this bad. This is 59 X 5 min subs taken over 2 nights (mainly last night). I took 250 new bias, 30 new darks and 70 new flats. I get the same result with or without the calibration frames only with dust motes if I don't add flats. There were no additional light sources last night compared to nights Ive taken my other images. Guiding was as normal and I was dithering every frame. Looking at the colour I thought it might have been a debayering problem but Ive gone through all the settings and they seem fine.


I've also added a picture of my best effort trying to process this crap so far. There is literally no colour in the galaxy at all. When I pull up anywhere in the saturation curve of through colour saturation it will increase colour everywhere like in the attached version.


Automatic background extraction wont get rid of it, nether will dynamic background extractor or colour calibration. I can only hide it.

I've included a dropbox link to the original untouched stack in case anyone is willing to have a go at this for me and shed some light on it.


https://www.dropbox.com/s/fzvw4njsrd...ched.xisf?dl=0

They're all drizzle integrations but I get the same result with the image files.

[Nikolas (Nik)]

looks like horizontal/vertical banding
Which camera were you using and what were the conditions temperature wise?

[Nikolas (Nik)]

also your file is .xisf and i dont use or want to use pixinsight

[Atmos (Colin)]

Are you dithering while taking the images?

Also, in PixInsight there is...
Script > Utilities > CanonBandingReduction

That might help.

[Gavin1234]

Using the ASI071MC. Sorry I uploaded the wrong file here is the TIF file. I took all the subs plus the calibration frames at -5c.

https://www.dropbox.com/s/45zpchvnjg...uched.tif?dl=0

Thanks Colin I just tried that utility but didn't seem to have any effect. No, only dithering after each exposure. I'm using SGP and PHD2 for guiding and capturing/sequencing.

Very strange, it seems like its striped all of the colour out of the stars and galaxy then dumped it into the background in those bands.

[Gavin1234]

I went back and restacked them for about the 50th time. This time I went back to blinking them and took out another 17 frames leaving me with 42. I thought I was being fairly ruthless already. The 17 I took out were just a slightly lighter shade of grey than then rest of them.

Anyway it made a pretty big difference. Either that or I repeatedly made the same mistake with one of the pre processing settings about 50 times in a row

Here is the latest version of it. Colour is probably a bit over the top as is usual for me I definitely need a lot more data for this target but at least I know I can actually process it when I get it now. Only 3.5 hours in this one, I think I would like about 35 hours.

[ChrisV (Chris)]

So what caused the banding Gavin? I'd like to know as I use the same camera. I'm trying to figure out what was in the bad subs - if that was it.

I've had blotchy backgrounds. But they were circles of blotchy and were due to using too small a scale when doing local normalisation.

[rcheshire (Rowland)]

Just a few thoughts - Given that the bands extend across the dithered stack, processing seems a likely cause. Difficult to say, if you changed processing options as well as removing frames in that last image.

It is also possible that the frames you removed did not respond well to your flats. Normalisation, if used, may exacerbate the problem.

As a base line, you could try processing a small image set (to speed things up) with a sample of the frames you removed, without normalisation (if you used it) and with bilinear deBayer (simple and usually safe) to see whether the issue reoccurs - as basic as possible.

[speach (Simon)]

Had the same problem, are you using Pixinsight? If you are when you debayer don't use the automatic setting instead select the bayer pattern that is appropriate for your camera

[billdan (Bill)]

Hi Gavin,
I went to download your TIF file to give Startools a crack, and then cancelled it when I saw it was 1.4Gb, that is one pretty big image for a stacked set of subs. You wouldn't have a FITS version of the stacked image would you?

I think I will stick to CCD cameras if that is the file sizes a CMOS stack produces.

[RyanJones]

Originally Posted by billdan

Hi Gavin,
I went to download your TIF file to give Startools a crack, and then cancelled it when I saw it was 1.4Gb, that is one pretty big image for a stacked set of subs. You wouldn't have a FITS version of the stacked image would you?

I think I will stick to CCD cameras if that is the file sizes a CMOS stack produces.

1.4Gb doesn't suprise me at all. Most of my stacks drizzled come to about that. I personally use custom frame stacking because I've got a full frame looking through a 1 1/4 apature and the vignetting is horrible so anything outside center is a waste anyway. That makes my stacks considerably smaller.

[Gavin1234]

Yeah I don’t really put any thought into making them smaller until I want to upload one to here. I’ll try shrink the stack, but the problem is sorted now anyway.

I know it looks like a debayering issue so I went through that and triple checked everything. My camera is RGGB and I always change to that from defaults.

I think you might be right about the normalisation. I use scale offset for my lights and equalise fluxes on flats. Although it was about 3am and my 50th odd attempt I’m fairly certain it was my usual settings. So I think it was the handful of frames I took out. I would bet it was some very high, very thin cloud that was messing things up. The frames I culled were slightly lighter than the rest.

I have about 30-40 hours of helix subs which I can’t process properly because I get similar issues when I add flats in, not the same banding but heaps of extra noise and blotching etc. I’m going to try that again just using only my darkest subs and then gradually add new batches in until I identify the bad ones.

Once again thanks for all the help guys. I would have thrown the towel in a few times by now if it wasn’t for this group.

[billdan (Bill)]

Thanks Ryan for your explanation, after your comment I later re-read that Gavin had drizzled his subs so that's why the stacked file is so large.

With my QHY12 OSC I'm used to a stacked image of approx 100Mb FITS file.

Cheers
Bill

[billdan (Bill)]

Hi Gavin,

With the Helix taking up a small portion of the full frame I wouldn't bother with flats unless you have heaps of dust bunnies. Crop out the wasted black space when post processing.

This last 12 months I have not done any flats and rely on the Startools Wipe module to fix up the vignetting and gradients (I keep the optics clean so I have no dust bunnies).

Cheers
Bill

[xelasnave]

I would try a restack telling ddss to crop to a size that removes the section at the top (and reframe to suit that removal ) them process that image...I think in startools you could remove it using "wipe" ..PI has something similar.
But why I say process without the bad bit is in startools at least that sort of thing can confuse it in effect.
Anyways look at all the good things in the image ..it is a distant gallaxy feel priveled to capture it☺
Alex

[xelasnave]

And a chance to enhance you photoshop skills.
Colour it in☺
My daughter did an Eagle Nebula for me using colours from a real flash image...
Cheating? Maybe.
Remove your stars and let startiol make new ones and rmeove all evidence of inappropriate camera settinngs and poor polar alignment...is that criket☺
The artist in me says do what you need to make you happy with the image.
A sort of marrying of the art of drawing with photography.
Alex