What are these odd oxygen patches?

[Datalord]

I'm processing an NB image using SHO and after combining I find these very odd patches of colour in the O3 data.

Top left is stacked O3, right is the combined SHO and the bottom is the raw sub.

They are all over image.

What is going on here?

Edited June 9, 2020 by Datalord

[CloudMagnet]

Not sure, could have been reflections or dust in the filter. Could rule that out if the patches dont show in each image. Do the patches show in each image or just some of them?

Were the O3 frames shot on the same night as the others?

[Datalord]

2 minutes ago, CloudMagnet said:

Not sure, could have been reflections or dust in the filter. Could rule that out if the patches dont show in each image. Do the patches show in each image or just some of them?

Were the O3 frames shot on the same night as the others?

They are in all frames, at the same places, stacking makes it worse. Agree, it could be something I might be able to sort with better flats.

Shot on the same night, yes.

[CloudMagnet]

Do they show in the flat frames as well? Would indicate dust or some obstruction.

Might be worth looking at previous nights images to see if these patches have ever shown up before- they might just have passed unoticed.

 

[Datalord]

I'm pretty sure it is a flats problem. I haven't taken new flats for my O3 filter for about a year, so better get a new set. I got around it this time by cloning these blobs out of the stacked image.

[Datalord]

The plot thickens. Like pea soup.

This is a sequence in blink. Same target, different times of night, patches come and go.

What?!

[IanL]

Reflection of an external light source that changes as the scope tracks?

[Datalord]

30 minutes ago, IanL said:

Reflection of an external light source that changes as the scope tracks?

Yeah, that's what I'm thinking. It's only present in O3 and at intervals. It's doing my head in.

[Datalord]

Nope, now it's showing on my Ha as well. Again, varying intensity in each sub, ranging from nothing to very visible.

[Datalord]

And, it follows the targets, so it is not something in the imaging train. If it was, it would be dithered out. I'm so confused... 😖

[CloudMagnet]

On 19/06/2020 at 17:46, Datalord said:

And, it follows the targets, so it is not something in the imaging train. If it was, it would be dithered out. I'm so confused... 😖

Hmm.. are you using a cooled camera? It might be patches on the sensor that are not being kept at the correct temperature throughout the night.

Does it show on any calibration pictures?

[Hallingskies]

Saw a similar thing on my ATIK 460, changed dessicant in desperation, problem went away.  Why bits of ice should migrate randomly across the chip is beyond me, but it seemed to be the problem...🤔🤔🤔

[Datalord]

3 hours ago, CloudMagnet said:

Hmm.. are you using a cooled camera? It might be patches on the sensor that are not being kept at the correct temperature throughout the night.

Does it show on any calibration pictures?

Yes, a G3-16200. It's rock solid -30C all night long.

11 minutes ago, Hallingskies said:

Saw a similar thing on my ATIK 460, changed dessicant in desperation, problem went away.  Why bits of ice should migrate randomly across the chip is beyond me, but it seemed to be the problem...🤔🤔🤔

hmm, interesting. But no, it can't be in the imaging train, as it retains its spot in the image, even with dithering. Meaning, it follows the target.

Edited June 24, 2020 by Datalord

[CloudMagnet]

Still been thinking about this. Here are some ideas:

1. Light reflections as Ian said. Easy way to find out is to point at a different part of the sky. If you still get the patern, we can safely rule this out.

2. Dew/moisture on scope. Possible as it would explain the change throughout the night but not likely as I can't imagine dew forming in those isolated clumps.

3. Filter related. Could be the brand of filters you are using has some internal reflections. Easy way to rule out is try a night without those filters in use.

4. Camera. Could be a cooling issue, might want to try some pictures just at ambient to see if the spots still show. If it is pixel related, it would show in calibration pictures, so I don't think this is to blame.

5. Aliens

 

Overall, it must be something hardware related to show in raw subs, so no processing to blame. Has this showed up in any nights imaging since?

[Datalord]

If I'm not mistaken, all of those issues, in the imaging train, would make the patches change position according to the guiding. It would be local to the "frame", not the "target". What I see in the subs is local to the target, that is what is mind boggling. 

In other words, I sort this out by doing a start alignment on all the subs, then clone stamp them away with the same stamp process across hundreds of subs. These things are local to the target. It's insane!