Misalignment of stacked images?

[Adam1234]

Hi, so I've taken some images of Markarian's Chain over 2 separate nights, but after I stacked the images in DSS and started processing in Photoshop, I noticed some weird artefacts - looked like small white dots or something - all over the image. I started processing again from the original stacked image and noticed the histogram looked weird - there were 2 peaks for each R G and B channel, which I'm assuming are related to these artefacts. Zooming in showed that they were present in the image before any processing. 

I also stacked the images from each night separately, and the individual nights stacked images and histograms looked fine, so I'm assuming it's some misalignment of the 2 nights images? I've never seen this before when stacking images from multiple nights. Any idea what's caused this?

I stacked the images from night 1 (11 x 300s, ISO400), night 2 (37 x 300s, ISO400) and both nights together with exactly the same DSS settings, and all images were taken with the Sky Watcher 80ED with Canon 2000d. 

The only difference between night 1 (8th May) and night 2 (14th May) I can think of is that the background sky in night 1's images were brighter than night 2's (I'll admit there was a bit of full moon on night 1). Would this be the cause? Thinking about it, I should probably just not use the images from night 1, the noise caused by the moonlight will probably counteract any benefit of more total exposure these 11 subs would bring but it would be nice to understand what's happening in the stacking as I plan on getting a few more hours on this target. 

The stacked image of both nights - zoomed in so you can see the white dots without me having to stretch the image

 

Stacked image of night 1 - again unstretched and zoomed in

 

Stacked image of night 2 - also unstretched in zoomed in

 

 

 

[MarkAR]

Looks like it didn't register the two nights properly. Do a little processing, identical to both the separate stacks to see if there is a difference in the moonlight ones.

If it's not too bad then you can manually align the two nights and blend into one. If it looks worse, then just use the second nights data.

[happy-kat]

In DSS did you put your light frames in separate groups for the different nights

Did you after registering set your best light frame as your reference frame

[Adam1234]

1 hour ago, MarkAR said:

Looks like it didn't register the two nights properly. Do a little processing, identical to both the separate stacks to see if there is a difference in the moonlight ones.

If it's not too bad then you can manually align the two nights and blend into one. If it looks worse, then just use the second nights data.

Thanks, will try that

[Adam1234]

1 hour ago, happy-kat said:

In DSS did you put your light frames in separate groups for the different nights

Did you after registering set your best light frame as your reference frame

Yes I put them in separate groups, but I let DSS choose the reference frame. What's the best way of determining the best light frame? Is that the frame with the highest score? If so I would have thought that would be what DSS chooses as the reference frame or is it slightly different when you use groups?

[happy-kat]

In your stack I would try a frame from the less Moon night as the reference. I choose on FWHM with score and unselect any particularly out of trend low scores.

[Adam1234]

2 hours ago, happy-kat said:

In your stack I would try a frame from the less Moon night as the reference. I choose on FWHM with score and unselect any particularly out of trend low scores.

I stacked again and chose a frame from the moonless night, with a score of +900. I don't think any of the frames were particularly low in score, they were all in the hundreds (I've had frames with a negative score before!) I still had the same thing. I'm probably just going to continue with the 2nd nights images.