How to stack Comets

I just did. Loaded the light and dark frames as I usually do. I had expected APP to not detect any stars at all given how bright the light-frames were, but much to my surprise it worked like a charm.

Note that these were taken with a DSLR and a 70-200 lens and not a telescope so there could be more data in them than what I thought.

Single 30s (tracked) Frame:

7 frames stacked in APP:

The comet is not even visible to the naked eye where I'm at because it is still so bright outside (59 degrees North)

First time using a tracker and was more of a test than anything else, so the mount-leveling was eyeballed and the polar alignment was very rough. Still (pleasantly) surprised of the outcome.

I think this works because the comet doesn't move very quickly. APP aligns subs on stars and not on the comet. I tried in the past and got very blurry comet pics with razor sharp stars.

Yep, Wouter is correct, APP only stacks on the stars. Which means the comet will keep on moving in the frames, making it more blurry. It only works nice when the movement is very minor. A comet stacking algorithm will require a bit of a redesign in the stacking algorithm, it is something we'd like to have as it's requested a lot, but will take a while.

@vincent-mod

Thank you for the explanation. Comet stacking is nothing exotic and should actually be available in a program like APP. I hope it will come soon.

@vincent-mod Why would this need a change in the stacking algorithm? Wouldn't it need a change in the registration algorithm only? In PI, you select the comet in the first and last pictures and PI interpolates the in between ones. I am not sure if it actually tries to find the center of the coma or if it simply assumes a linear motion with steps based on the time difference between the subs, but I think that's all there is to it. In a first order approach, the assumption of linear motion will be more than sufficient given the relatively small time ranges over which comet images usually are taken. Still, it will require time to implement this of course.

Thanks, Wouter

You’re right Wouter, it probably isn’t really the stacking algorithm itself that needs a lot of work, but it’s still a module that needs to be rewritten (that or the registration one) or added to it this possibility. Or another module all together. I have no experience with these algorithms myself so how difficult this is or not, I have no idea. I guess it’s not rocket science for Mabula. 😉 I know it’s on the to-do list, but things like noise reduction etc are higher I think.

@kpsebsdorf Yes it will, but as with all software development paths... it’ll take time. APP first and foremost is a deep sky stacking program and tries to be top-notch in that, not a solar object processor. Although these possibilities will be added in the future.

Well well, when you're talking about it it of course is going to change... stay tuned. 🙂

Here's an image of a comet that is stacked with alignment on the comet head - using Nebulosity. You can see that the result is probably not what you're looking for - with stars streaking behind the comet. 

This comet, Neowise being such a great comet doesn't need stacking or you can stack on stars for 15 mins or so. Next is a single 10s image that could easily have been longer, like minutes, without losing detail or showing star trails.

The following is an APP stack (somewhere around 20-40 10s images 200mm lens)

Here are a couple of excellent nightscapes - stacks of only 2 images - on the stars no doubt. I'm no expert but I question the need to have a special stacking algorithm for comets, unless maybe you are trying to do some kind of science.

https://www.pentaxforums.com/forums/groups/135-astrophotography/9984-neowise-its-best.html

A couple days back I captured 15 each of ISO400 and ISO800 using 1.5 minute shutter speed.

There is a leading edge distortion when stacked which I don't understand as this isn't in the individual frames.

Would there be a preferred method to not end up with the leading edge distortion?

@paul-from-northern-mi

Your total capture period is 2 x 1.5 x 15 = 45 minutes so the comet has moved that far - the length of the head - in the star field.

If you look at my stack just before, I had many exposures but they were short so that the total capture duration was only about 15 minutes - I forget the exact amount I stacked because I collected 200 but only stacked a fraction of them so that there wasn't too much movement in the comet.

Posted by: @astrogee

@paul-from-northern-mi

Your total capture period is 2 x 1.5 x 15 = 45 minutes so the comet has moved that far - the length of the head - in the star field.

If you look at my stack just before, I had many exposures but they were short so that the total capture duration was only about 15 minutes - I forget the exact amount I stacked because I collected 200 but only stacked a fraction of them so that there wasn't too much movement in the comet.

I do see what you are talking about.  The movement of a comet would not match with the stars.

I would have thought though that the movement mismatch would be front to back instead of side to side.  Also the distortion on both sides of the front is symmetrical to the center of the tip of the comet.

I did see someone on Youtube who used 2 minute subs.  I stepped my frames down slightly from that.  I might want to look closely at his results.

@paul-from-northern-mi

I'm not a comet physicist but the comet direction is side to side. The tail is in the direction of the sun as the solar wind is what pushes the tail out.

If you look at my first image, you can see that it is a stack registered on the comet head. The stars trail side-to-side showing that the comet is moving side-to-side.

Posted by: @astrogee

@paul-from-northern-mi

I'm not a comet physicist but the comet direction is side to side. The tail is in the direction of the sun as the solar wind is what pushes the tail out.

If you look at my first image, you can see that it is a stack registered on the comet head. The stars trail side-to-side showing that the comet is moving side-to-side.

You are very correct.  In briefly viewing this  

I can see that Peter Zelinka had the same issue as I in stacking with longer exposures.  I haven't looked at how he solved the stacking issue but I believe that it likely was done completely in Photoshop.

Hello,

can we have news and links here about the new comet processing coming up in APP as announced yesterday in facebook?

stefan

A public beta version will become available very soon. I expect within a week.

Any news on this? Got a nice stack of images of Leonard with M3......

@marco_verstraaten Comet processing has been part of APP since version 1.83-beta1.

Stars and comet stacked separately and blended? That is what you want with M3 in the image!

@marco_verstraaten Ah you didn't mention that in your question. No, that not yet.

@wvreeven I second @marco_verstraaten's request to have the stars and comet both aligned in the same image. Even Deep Sky Stacker has this capability. I am on Mac otherwise I would use that. But currently I have no way of stacking them both without doing each individually, then removing the stars in the comet stack (which gets messy), and then combining them later. Would much rather have an easy comet + stars stacking feature in APP!

Edit: I just watched Mabula's tutorial video on comet stacking in v1.083 beta and at the very end, around the 19-minute mark, he very helpfully explains how to reject stars in the comet stack and it worked great. Still, though, I think this feature would be extremely helpful!

@absorbingphotons I third the request.  🙂  Here's 60 minutes of L data with Leonard's conjunction with M3.  I'd love to use APP to make the stars/M3 pop!  I haven't figured out how to do it in Pixinsight yet as I would prefer to just do all of my pre-processing in APP.

@wvreeven Ooh, nice! Looking forward to trying it. I'm headed out now to try my luck at shooting Leonard.

I'm still on 1.082, though, and when I look at my Downloads page under My Account, I see nothing but 1.082. How can I find the new new beta? Hopefully it doesn't require a new purchase?

@rdviii on this page, look at the top right. there's the download link. APP dowload & release notes.

@marco_verstraaten Awesome, thanks.

My first Leonard comet-registered stack. 21x40 seconds, Redcat 51 (250mm f.l., f/4.9). Sony a7iii @ISO1600, 4.94 arcsec pixels. Subs span about 21 minutes, downselected from 31 for tracking problems.

I'd say it came out okay! Was hoping to catch more than the 3-4 degrees of tail. I have about 4x as much data on hand, taken with different framing, will see what it looks like when I max that out.