Error with Comet stacking (stars trailing)

[Bagginsies]

Hey all,

I have been using Deep Sky Stacker to stack my comet images. I have used the stack stars and comet before without problem, but for some reason this past few months, every time I stack now I get a slight star trailing as in the image below.

I decided to have a quick go at Y1 Atlas last night (knowing the Moon and weather were not on my side) just to have some data to play with and test, but ended up with the same star trailing.  

I don't know what could be causing this as all my light frames (60sec @ iso1600) are clean, when I stack it as a normal star stack the stars are pin sharp (comet blurs obviously, but this tells me that the tracking is fine), I have made sure the comet selection is good checking each one thoroughly, I have even reset to default settings. I used to just click, save and stack and it worked fine. Not sure what is going on now?

 

20200505 C2019 Y1 Atlas.TIF

[Bagginsies]

An edited single image

[Bagginsies]

Forgot, top image with trailing I didn't use flats or bias frames (only darks), but even when I use flats and bias I still get trailing.

[vlaiv]

Speed of the comet combined with exposure length?

Comet stacking works by rejecting certain pixels from the stack.

You can either align subs on comet - stars will trail or on stars - comet will trail. Comet stacking works by doing both - taking image of the comet from stack aligned on comet and stars from stack aligned on stars. Issue is to distinguish what are stars and what is comet - and it does that by examining the pixels in each stack - how much they change between subs.

Sometimes, if comet is too slow - there is not much change between star positions in the image and algorithm can tell if that is something to do with comet (it stayed the same with respect to comet) or is it star (moved enough with respect to comet) - so you get above mess.

Not sure how you can try to make it better - are there any parameters for stacking that you can tweak?

[Bagginsies]

When I look at each frame there is movement in the stars and the comet is slow so that could be a problem for the software as you say above. 

To be honest I had always just used the standard settings and not sure what I could/should tweak as it would just be guessing. Will have to look into it more, so much to learn and get my head around LOL. I was hoping someone would say "just click this button and hey presto" ha ha, if only!

I have also just seen another thread on here about doing the two stacks (one on comet and one on stars) separately in DSS then blending myself in my photo editing software (Affinity Photo), again, so much to learn!

….but I know it will all be worth it in the end 🙂

[happy-kat]

Using the comet feature I would mark the comet on each frame, when stacking I would stack on comet and stars and use kappa sigma clipping. Perhaps that will help.

Did you use dithering.

[carastro]

I could never get the comet stacking feature on DSS to work properly, it would stack some OK, but not others, and after messing around for ages I gave up and stacked them manually in Photoshop.  I stacked so the comet was stationary and the stars trailed. 

Carole 

[Bagginsies]

1 hour ago, happy-kat said:

Using the comet feature I would mark the comet on each frame, when stacking I would stack on comet and stars and use kappa sigma clipping. Perhaps that will help.

Did you use dithering.

I have used those settings and it didn't work (don't really understand it all yet, so just checking boxes half the time!). As for dithering, I don't think the box was checked, but that is something I have yet to understand/test. The learning continues...….. LOL

1 hour ago, carastro said:

I could never get the comet stacking feature on DSS to work properly, it would stack some OK, but not others, and after messing around for ages I gave up and stacked them manually in Photoshop.  I stacked so the comet was stationary and the stars trailed. 

Carole 

I had the same with some working and some not, will try other methods to see what I can do, but like vlaiv pointed out, I think some better data in my light frames will help DSS in it's job.

I had a great picture a while back, but could only stack on the comet, luckily the picture still looked good like that. I will use the next few weeks to try out some manual stacking etc. At least I will have something to do for the cloudy nights 🙂

[carastro]

Did you take flats?

Carole 

[Bagginsies]

27 minutes ago, carastro said:

Did you take flats?

Carole 

The top image didn't have flats, but even when added, it still comes out the same. 

I am thinking it is the software struggling to keep the comet and star positions right when stacking. When you go through the light frames, the stars jump around a lot. Although the pictures are clear, the stars move a bit each time (not in the same direction either so not tracking, it would be like taking a pictures hand held), so could be tiny movement from a loose gear or some other tiny movement in the setup. When stacking this then makes it hard for the software to keep up with position of everything in the light frames and this makes it put one or two frames in a different location thus creating the lines....if you get what I mean, not sure I even do LOL

[Bagginsies]

Just tried a new stack with kappa-sigma clipping and this I took each light frame and increased the clarity (made stars and comet brighter) of each light frame. The final stack was better but had a double image. i.e. every star had another below it. Looking at it I may have even forgot tot stack the comet! Think I need to wait and get some good data and try again.

Edited May 6, 2020 by Bagginsies

[Bagginsies]

I have re-installed DSS and took some images of M13 last night (10x90sec) I also spent some time looking at the backlash settings as they were fairly high, I have reduced them and it seems smoother. Not even sure if it makes a difference in actual tracking? I now know it isn't the tracking as M13 was fine.  Image was stacked in DSS and turned out well. Will wait for a really clear night and try again on a comet.