comet registration and star trail removal

Yes, I am having a similar issue, either sharp comet with star trails using the comet registration option, or sharp stars with moving comet using standard star registration.  I recall seeing on the forum in July 2022 that there was to be an update to achieve both sharp comet and star background.  Is there a planned update coming for this?

thanks!

It is still on Mabula's list yes, since we had to change a lot in our development platform and had to prioritize other features, it's still in progress.

@vincent-mod would you happen to know the priority on finishing the comet stacking feature so we can have have the sharp comet / sharp stars? I have been collecting comet data over the years that I can't do much with since I only use APP for all my processing. I hate to be part of the crowd that keeps asking because I know there is a lot of hard work going into this. But some info on what to expect would be nice to know if this is something to look for soon, or if it is put on the shelf until other projects are finished first?

Unfortunately I don't know the schedule Mabula has, I did forward the requests as of course it's comet-season. 🙂 I'll put forward that this is a feature we should put higher on the priority list.

@vincent-mod I definitely don't want to sound like I know the difficultly of authoring programs like this because I don't. But is there a programing challenge to overcome to make this available for us? Or is there just too many other things to complete first before this is worked on? It seems that since we can already align on the comet and also align on the stars depending on the setting that is selected, all we would need to a 2 stack process to complete a 100% sharp comet image. Pretty much how DSS does it. That is just my thinking, but like I said, I am no programmer...

Hi Damian @dtc1999,

I will move this feature upwards now in our development path 😉 I will try to work on this in the coming period. Like Vincent indicated, we needed to fix a lot of issues with APP 2.0.0 on different platforms, and all of these issues are now fixed we think and APP is ready going forward ! I will check how DSS performs on this as well.

Mabula

Couldn’t people use a star removal program in the meantime like Starnet++ or starXterminator or just masking in Affinity or PS or Gimp and then combine the two results?  (I.e the sharp stars with the sharp comet?).  I know it would be better to have APP do everything, but there is no free lunch and I’m wondering if it would be better to focus on APP features that can’t be easily replicated somewhere else as the top priorities.

I combine RGB stars with narrowband nebula using this method. I would think this would be just part of the basic AP toolbox for most to combine the star layer from one image with the nebula from another.  I know not everyone has the software to do this, but I strongly encourage you to learn if you don’t.

@mabula-admin Thank you Mabula. I look forward to this and everything that comes next for APP. 

You are most welcome Damian @dtc1999

I'm not too bothered about having an single process to get both a sharp comet and sharp stars in one go, but I'm struggling to get a comet image without stars or star trails in it.

I've done the comet registration and that works just fine with a great comet, but I'm getting blobby star trail left overs. This is despite having a long integration time AND selecting the median option.

Given stars are not in the same location or overlapping locations for anything like half the time with my data, I would have expected median to kill them, as the median star value in any pixel is zero. So why are am seeing star trails at all? I'm using 1.083 beta1 - maybe that is a bit old, but I'm baffled as to what "median" is doing now. If there is a collection of options I can set with or without an update to 2.x beta y then please let me know that too. 

Further to my previous post:

1. I have upgraded to 2.0.0beta13 and that does not seem to have improved the situation

2. I have experimented meantime with the clipping settings. When I first posted I had forgotten to turn off clipping, which could obviously mess up the median as it appears clipping is applied first, but doing a pure median, while better, still has blobby star trails. 

3. While the use of the median (at least for larger stacks) might be sub-optimal from an overall image quality point of view, and a simple/clipped average (but with an intelligent algo for removing the trails) would be preferable, median stacking really ought to work here. I'll keep rummaging around the settings but it really feels like "median" is not doing the operation it ought to. The trails of my stars are rather long and they are spending a small fraction of the overall integration time in any one place. So the pixels are bright from a star presence well less than half the time. So median should kill the star trail. 

I'll keep fiddling with the settings to see if I have interfered with the operation of median inadvertently somewhere. I also wondered if the implementation in APP was not a pure median, perhaps to remove potential posterising of faint details?

The Image shows the effect of comet registration followed by median stack without clipping, then a background extraction and stretch. (The posterising around the comet head is JPG artefact)