Another excellent tutorial. I've taken several mosaics now but have always just settled for the default result. I'll give this one a go and compare. Thank you, your effort in teaching is always very much appreciated.
Thanks for the step by step. After watching your videos, Santa is bringing an s50 to our house, and so we are about to get our first taste of astrophotography
I don't have my first telescope yet. I hope it will be an S50, however I enjoy every video you make because you take a subject that, in my ignorance, seems difficult and complex and you make it look easy. Thank you and clear skies.
@@ElkinGuillermoForeroRosado seestar s50/s30 is actually very easy and simple to use 😊 only challenge is the weather 😬 And get planetary focus. I'm still struggling to get a better focus on jupiter and Saturn than my camera can get.
Dear Madam, Thank you very much for a bunch of very "good-to-know" informations, As soon as I get enough subs of the western veil nebula (mosaic mode aswell), I'll give it a go ! Cheers from sunny south-east of France ! Merci !
Many thanks for that, it really helped a lot. I was fighting with visible steps between individual mosaic pieces. After I edited everything again with your instructions, these have now completely disappeared! Thank you very much!
@@the_space_koala I run PixInsight on a MacBook Pro with an M3 Pro processor. 542 lights only took about 45 minutes to complete (debay, alignment and integration). My 'old' MacMini with a M2 Pro needs about 1.5 hours. So, a fast CPU is really what you need for stuff like that. :o)
You need many cores and a lot of memory! I ran this on my Mac Studio with a M1 Ultra and 64GB unified memory. Also for me it didn’t take very long. But I know that the average viewer will have a longer run time and I want to manage expectations 😁
Another informative and understandable tutorial from you. The additional information about the settings in Star Detection and Star Matching was especially helpful - Thanks soooo much!!!! I've used this approach (based on your previous tutorial) and, because of my high light pollution and/or next door neighbor's lights, I found that some of the "edges" between overlapping frames showed up in the final stacked image making it unusable. I tried applying Automatic Background Extraction to all the Debayered images before doing the ImageIntegration and it solved the problem. I don't know if this is useful for everyone, but, it has allowed me to get good stacked mosaics for several targets. Just FYI. Steve
that's a great idea on the ABE. I did a version of this video where I also used flats but I feel my flats strategy doesn't provide enough benefits for the effort it takes :D
Hey Luca, great video, very informative. Set up the process on Pixinsight and it worked a treat on a mosaic of M42 with only 300 10s subs and with a bit of cropping of the final image was pretty good. Keep up the good work.
Wow. Pixinsight is such a powerful tool. I don't have it, yet, but as my first year in this hobby approaches, I often get the feeling of how or what can I do to improve the image. Ten months ago, I never even heard of post. My goal is to get a dedicated PC just for processing and yes, getting and learning Pixinsight. Your video is very informative, but would be easier to follow along if I had the SW. Still, good job, thank you.
Pixisnight is undoubtedly a huge learning curve but I’m convinced it is the ultimate tool for Astro processing. I purchased it about a year after starting Astro and I wish I’d done it earlier.
Thank you very much for this excellent tutorial. I wonder if you would consider showing the full PixInsight process in one of your next videos? What I mean is the use of Cosmetic Correction (with CFA), Debayering, Weighted Batch Processing (do I really need this for Seestar images?), Star Alignment, Image Integration, Dynamic Background Extraction (DBE), and Spectrophotometric Color Calibration (SPCC) for mosaics as well as single frame stacking. PixInsight seems to be the top dog, but the learning curve is enormous. Your shortcut process was VERY helpful for me. Thank you! I got my Seestar S50 two weeks ago (first telescope) and I am very impressed what is possible today to get amazing results in no time.
This is great! Thanks for the tutorial. I have done a "master" mosaic attempt for M31across multiple nights. The S50 can do deep sky stacking (manually, not live) on the S50 device itself over multiple days of data... so I will try to use that output as a reference image in PixInsight when my trial is approved.
Sorry for the delay i needed a second nights shooting to test out the multiple night shooting and the answer is yes. The manual stacking option offered both nights shots for selection and combined them with out any problem.
Thanks - some excellent thoughts and valuable insights here. After a recent firmware update, the Seestar has become much capable at restacking frames into mosaics. The frames can be from mosaics or regular imaging. If they are in a folder on the Seestar ending "_mosaic_sub", then you can ask it to restack and it will make a big mosaic out of them. Probably not as nicely as with the method you describe, but with no effort on the part of the user (other than moving files around). I have found that c. 2700 frames takes about 2.5 hours and needs about 30 GB of free space on the Seestar. I have recently put some "Mega-Mosaic" images on my astroimg account (with the same username, 64gkb, as here). The image tends to have messy edges that need cropping away, and can be an odd shape depending on how the images stack up.
Hey Luca, great video! I've worked out a way to register and integrate mosaic images using WBPP while using the entire field of view t hat the Seestar usually crops out. I posted that video a few weeks ago if you're curious. Clear skies!
@@the_space_koala Yeah I agree, they need to open up some of the parameters to the UI. My method involves inputting those customizations in the code and once they're in, WBPP has almost no issues with registration. The largest dataset it was tested with was about 6500 frames I believe. And I'm currently trying to gather a giant mosaic of the M81/M82 area to test. In the near future, I think Siril will beat Pixinsight for Seestar Mosaic stacking. I also have a demo of that and Siril registers and stacks much faster than PI. Once that's in Production, my workflow will probably change to using Siril for registration/stacking and PI for post-processing. As much as I love PI, WBPP is just way too slow, even when you turn off drizzling and weighting.
I agree on this, siril is ridiculously fast. Also because I use a Mac, Siril is optimized to use the ARM chip while Pixinsight still relies on translation to compile which I believe takes away from the efficiency. That said, Siril is just not flexible enough for me so I'll keep my whole workflow in PI, it's not slow enough for me to think about switching
Thank you some much for this guidance! It appears that the Registration Model Options for the StarAlignment tool have changed recently, possibly in the new Lockhart update of PixInsight. I no longer see the option for choosing "Thin Plate Splines" as describe around the 16:00 in your tutorial; however, one can select "Two Dimensional Surface Splines." Doing so opens up the option for selecting "Thin Plate Spline" as the "Radial Basis Function," which I don't recall seeing previously. Do you think this is a good path forward?
yess they have changed the ImageIntegration process completely. But you're correct - that's the new equivalent. I wish I could add a note on the video for others
@@the_space_koala Thank you for the prompt response. Last night I experimented with the above adaptation of your process using a few hundred frames of NGC 869 and NGC 884. The result was quite good, super-accurate but super-slow, maybe because most of the frames had over 1000 star pairs. Thank you again. I'm a lowly chemist by training but am really appreciating what are you doing for people in this community.
Super useful thank you. I'm trying to process my first mosaic, Ghost of Cassiopeia, and was getting criss-crossing lines in my stacked image where the edges of the images overlapped. I don't see them anymore after using your method, perhaps because the weights were set to "don't care"? Thanks all the same.
Hey, great video! I currently do not have Pixinsight, and use Siril for processing. Do you think this same method can be used in Siril? I can imagine there's a similar workflow to accomplish the same result, I'm just curious on your thoughts since they are different programs at the end of the day. Thanks!
Siril can do a lot of the things pixisnight can - and it does it much faster too - but I’m afraid there are much fewer customization options for the registration step. Otherwise you’d follow the same exact process. Can you specify a reference image that is not part of the series though?
Yes. For all of us just starting out, investing in Pixinsight right off the bat isn’t such a good choice. Doing this on free software would be a lot more helpful.
I would suggest Asto Pixel Processor. Undoubtedly the very best at mosaics and much faster, easier and cheaper than Pixinsight. Siril will have the ability to do complex mosaics in the near future.
We have had snow multiple times this year but it melted each time near my house - an hour north from me in the proper mountains roads have been closed for a month due to snow though ❄️
This backward engineering of the Bayer mask was valuable information. Thanks for that. Have some pictures where the colors just didn’t look right. I’m gonna process em one more time experimenting with the Bayer pattern.
Particularly useful since low Earth orbit satellites are most visible early during the night when they are still illuminated by the sun. So in general they are more likely to appear in the first few sub frames when there isn't any data to do automatic rejection yet.
Thank you so much for your step by step tutorial. I'm shooting my first mosaic right now, but i only got 300 raws during 3.5h out of the swiss sky... How do you get so much more out? Is there any secret to it? Kind regards
@the_space_koala thank you for your answer. I'm aware that multiple nights are needed. It's just that with the rejection rate i'm experiencing i would need roughly 50hrs to get 2700 images... How many hours sis you shoot?
I think that you can maintain "projective transformation" in "Registration model", and it is better check in "Rigid linear transformations", because there are no deformation.
Great vid thanks! Good explanation of why to do it! Hadn’t appreciated all the settings in star alignment so thanks for that! Is fast integration useful to speed things up a bit? But loses quality? Also live stacking cannot do the frame weighting correctly as it doesn’t know what’s to come. Thanks!
I don't have pixinsight, yet, but I'm collecting mosiac data as an incentive to pick it up, soon. I have a really good completed frame of the east veil and Pickering's triangle, and I'd like to use it as a reference for future shots, but I forgot to screencap my framing. Is there any way to calculate this so that I can collect the same data for future nights, or am I going to have to guestimate?
@the_space_koala an eclipse normally take a few hours, that's why I'm worried. I've only tried single shots both with my camera and the seestar. I have 15 stop nd filter for the camera.
Hi whilst this is an exellent tutorial i have found that this teqnique creates ugly banding on my images is thare any reason for this? I have poseted a more detailed explanation of this on the Pixinsight forum and referenced Space Koala thanks in advance
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
Another excellent tutorial. I've taken several mosaics now but have always just settled for the default result. I'll give this one a go and compare. Thank you, your effort in teaching is always very much appreciated.
Thanks! Do let me know if you get better results
Thanks for the step by step. After watching your videos, Santa is bringing an s50 to our house, and so we are about to get our first taste of astrophotography
Oooh, congrats! Clear skies
Just got mine. So fun!
I don't have my first telescope yet. I hope it will be an S50, however I enjoy every video you make because you take a subject that, in my ignorance, seems difficult and complex and you make it look easy. Thank you and clear skies.
That's really kind of you to say! Hope you get your S50 soon
@@ElkinGuillermoForeroRosado seestar s50/s30 is actually very easy and simple to use 😊 only challenge is the weather 😬
And get planetary focus. I'm still struggling to get a better focus on jupiter and Saturn than my camera can get.
Dear Madam,
Thank you very much for a bunch of very "good-to-know" informations,
As soon as I get enough subs of the western veil nebula (mosaic mode aswell), I'll give it a go !
Cheers from sunny south-east of France !
Merci !
Many thanks for that, it really helped a lot. I was fighting with visible steps between individual mosaic pieces. After I edited everything again with your instructions, these have now completely disappeared! Thank you very much!
I’m glad this helped! You were quick to reprocess this thing takes forever 😁
@@the_space_koala I run PixInsight on a MacBook Pro with an M3 Pro processor. 542 lights only took about 45 minutes to complete (debay, alignment and integration). My 'old' MacMini with a M2 Pro needs about 1.5 hours. So, a fast CPU is really what you need for stuff like that. :o)
You need many cores and a lot of memory! I ran this on my Mac Studio with a M1 Ultra and 64GB unified memory. Also for me it didn’t take very long. But I know that the average viewer will have a longer run time and I want to manage expectations 😁
Absolutely great video. I have learned so much for processing this Seestar images. ❤ Please more of that. Thx.
Thank you! Will do!
Once again a video that explains a subject in a normal understandable language 🙂. Keep up the good work.
Thank you!
Another informative and understandable tutorial from you. The additional information about the settings in Star Detection and Star Matching was especially helpful - Thanks soooo much!!!!
I've used this approach (based on your previous tutorial) and, because of my high light pollution and/or next door neighbor's lights, I found that some of the "edges" between overlapping frames showed up in the final stacked image making it unusable. I tried applying Automatic Background Extraction to all the Debayered images before doing the ImageIntegration and it solved the problem. I don't know if this is useful for everyone, but, it has allowed me to get good stacked mosaics for several targets. Just FYI.
Steve
that's a great idea on the ABE. I did a version of this video where I also used flats but I feel my flats strategy doesn't provide enough benefits for the effort it takes :D
Hey Luca, great video, very informative. Set up the process on Pixinsight and it worked a treat on a mosaic of M42 with only 300 10s subs and with a bit of cropping of the final image was pretty good. Keep up the good work.
glad this worked! Happy holidays
Thanks for the awesome explanation on of how software decides what pics to discard and keep, very informative! Thanks!
I’m glad the explanation was clear! Thanks for the feedback
Wow. Pixinsight is such a powerful tool. I don't have it, yet, but as my first year in this hobby approaches, I often get the feeling of how or what can I do to improve the image. Ten months ago, I never even heard of post. My goal is to get a dedicated PC just for processing and yes, getting and learning Pixinsight. Your video is very informative, but would be easier to follow along if I had the SW. Still, good job, thank you.
Pixisnight is undoubtedly a huge learning curve but I’m convinced it is the ultimate tool for Astro processing. I purchased it about a year after starting Astro and I wish I’d done it earlier.
Thank you very much for this excellent tutorial. I wonder if you would consider showing the full PixInsight process in one of your next videos? What I mean is the use of Cosmetic Correction (with CFA), Debayering, Weighted Batch Processing (do I really need this for Seestar images?), Star Alignment, Image Integration, Dynamic Background Extraction (DBE), and Spectrophotometric Color Calibration (SPCC) for mosaics as well as single frame stacking. PixInsight seems to be the top dog, but the learning curve is enormous. Your shortcut process was VERY helpful for me. Thank you! I got my Seestar S50 two weeks ago (first telescope) and I am very impressed what is possible today to get amazing results in no time.
yess, an end-to-end processing video is coming in the next weeks
thanks for the video🥰
This is great! Thanks for the tutorial. I have done a "master" mosaic attempt for M31across multiple nights. The S50 can do deep sky stacking (manually, not live) on the S50 device itself over multiple days of data... so I will try to use that output as a reference image in PixInsight when my trial is approved.
I think the trial license is approved very quickly! Good luck
Well done! Very clear tutorial.
Thank you. I can't wait to try this out.
Thank you very much. So informative and clear.
You can restack mosaics in the Seestar and that allows you to check the subs and manually exclude the ones with satellite trails.
if you restack, you shouldn't need to exclude satellite trails as it should reject those pixels. Does it let you stack multiple nights?
Sorry for the delay i needed a second nights shooting to test out the multiple night shooting and the answer is yes. The manual stacking option offered both nights shots for selection and combined them with out any problem.
thank you. this tutorial is great
Thanks - some excellent thoughts and valuable insights here.
After a recent firmware update, the Seestar has become much capable at restacking frames into mosaics. The frames can be from mosaics or regular imaging. If they are in a folder on the Seestar ending "_mosaic_sub", then you can ask it to restack and it will make a big mosaic out of them. Probably not as nicely as with the method you describe, but with no effort on the part of the user (other than moving files around). I have found that c. 2700 frames takes about 2.5 hours and needs about 30 GB of free space on the Seestar. I have recently put some "Mega-Mosaic" images on my astroimg account (with the same username, 64gkb, as here). The image tends to have messy edges that need cropping away, and can be an odd shape depending on how the images stack up.
does it let you combine multiple days of mosaics?
Wow…thank you❤
Really awesome video, thank you. If I wanted to drizzle my image, how would I integrate that process?
Hey Luca, great video! I've worked out a way to register and integrate mosaic images using WBPP while using the entire field of view t hat the Seestar usually crops out. I posted that video a few weeks ago if you're curious.
Clear skies!
Thanks I’ll check that out! I just feel WBPP just doesn’t offer enough customization options for the registration to get all frames done
@@the_space_koala Yeah I agree, they need to open up some of the parameters to the UI. My method involves inputting those customizations in the code and once they're in, WBPP has almost no issues with registration. The largest dataset it was tested with was about 6500 frames I believe. And I'm currently trying to gather a giant mosaic of the M81/M82 area to test.
In the near future, I think Siril will beat Pixinsight for Seestar Mosaic stacking. I also have a demo of that and Siril registers and stacks much faster than PI. Once that's in Production, my workflow will probably change to using Siril for registration/stacking and PI for post-processing. As much as I love PI, WBPP is just way too slow, even when you turn off drizzling and weighting.
I agree on this, siril is ridiculously fast. Also because I use a Mac, Siril is optimized to use the ARM chip while Pixinsight still relies on translation to compile which I believe takes away from the efficiency. That said, Siril is just not flexible enough for me so I'll keep my whole workflow in PI, it's not slow enough for me to think about switching
Still waiting for my S30 to be shipped after having waited two weeks now. Ordered it after watching you track the airplane.
I hope you get it before Christmas!
I pre-ordered in November. Latest update says delivery late December or early January
@@troyjhowlettdid you order directly from zwo? I havent seen any updates from them
I did the same thing, but have cancelled it because ZWO has given a new time line. It at least February before you get it.
@@willrothfuss8470 where did you see this? Their website still shows late december
Thank you some much for this guidance! It appears that the Registration Model Options for the StarAlignment tool have changed recently, possibly in the new Lockhart update of PixInsight. I no longer see the option for choosing "Thin Plate Splines" as describe around the 16:00 in your tutorial; however, one can select "Two Dimensional Surface Splines." Doing so opens up the option for selecting "Thin Plate Spline" as the "Radial Basis Function," which I don't recall seeing previously. Do you think this is a good path forward?
yess they have changed the ImageIntegration process completely. But you're correct - that's the new equivalent. I wish I could add a note on the video for others
@@the_space_koala Thank you for the prompt response. Last night I experimented with the above adaptation of your process using a few hundred frames of NGC 869 and NGC 884. The result was quite good, super-accurate but super-slow, maybe because most of the frames had over 1000 star pairs. Thank you again. I'm a lowly chemist by training but am really appreciating what are you doing for people in this community.
@@geneehandley9902 actually I am also a lowly chemist by training (medicinal) ;) though I don't work in my field
Thankyou! I appreciate the examples! The example image from minute 4:00 was taken with seestar s50?
@@RobAguilarPhoto yes that’s the Andromeda mosaic I did (it’s shown in more detail in my original mosaic video)
Super useful thank you. I'm trying to process my first mosaic, Ghost of Cassiopeia, and was getting criss-crossing lines in my stacked image where the edges of the images overlapped. I don't see them anymore after using your method, perhaps because the weights were set to "don't care"? Thanks all the same.
yes I think it's the weighting that makes the difference. glad it helped your mosaic
I am waiting for a nice beginners tutorial on pixinsight from you!
It’s one of the (many) things in the pipeline, but it’s not too trivial if I want to do it well :)
Hey, great video! I currently do not have Pixinsight, and use Siril for processing. Do you think this same method can be used in Siril? I can imagine there's a similar workflow to accomplish the same result, I'm just curious on your thoughts since they are different programs at the end of the day.
Thanks!
Siril can do a lot of the things pixisnight can - and it does it much faster too - but I’m afraid there are much fewer customization options for the registration step. Otherwise you’d follow the same exact process. Can you specify a reference image that is not part of the series though?
Yes. For all of us just starting out, investing in Pixinsight right off the bat isn’t such a good choice. Doing this on free software would be a lot more helpful.
I would suggest Asto Pixel Processor. Undoubtedly the very best at mosaics and much faster, easier and cheaper than Pixinsight. Siril will have the ability to do complex mosaics in the near future.
Need more images of your beautiful mountains. Any snow yet?
We have had snow multiple times this year but it melted each time near my house - an hour north from me in the proper mountains roads have been closed for a month due to snow though ❄️
This backward engineering of the Bayer mask was valuable information. Thanks for that. Have some pictures where the colors just didn’t look right. I’m gonna process em one more time experimenting with the Bayer pattern.
If you don’t have the weird pattern just the wrong by colors it should be enough to swap the red and blue channels on the final image
@ gonna try this first. Fits better to my „order of the lazy geek“ approach ;o)
Particularly useful since low Earth orbit satellites are most visible early during the night when they are still illuminated by the sun. So in general they are more likely to appear in the first few sub frames when there isn't any data to do automatic rejection yet.
Yeah it’s the worst in the summer when they’re visible all night 😁
Thank you so much for your step by step tutorial. I'm shooting my first mosaic right now, but i only got 300 raws during 3.5h out of the swiss sky... How do you get so much more out? Is there any secret to it? Kind regards
you truly need to do multiple nights, otherwise it's just impossible to get enough data for a mosaic, sadly
@the_space_koala thank you for your answer. I'm aware that multiple nights are needed. It's just that with the rejection rate i'm experiencing i would need roughly 50hrs to get 2700 images... How many hours sis you shoot?
I think that you can maintain "projective transformation" in "Registration model", and it is better check in "Rigid linear transformations", because there are no deformation.
There could be deformation due to imperfect optics and as they’re not 100% aligned it helps if you do thin plate splines
@the_space_koala It's true. There is air turbulation, too. Thank you for your answer.
Great vid thanks! Good explanation of why to do it! Hadn’t appreciated all the settings in star alignment so thanks for that! Is fast integration useful to speed things up a bit? But loses quality? Also live stacking cannot do the frame weighting correctly as it doesn’t know what’s to come. Thanks!
I don’t think fast integration is an option here because it uses its own registration process
I don't have pixinsight, yet, but I'm collecting mosiac data as an incentive to pick it up, soon. I have a really good completed frame of the east veil and Pickering's triangle, and I'd like to use it as a reference for future shots, but I forgot to screencap my framing. Is there any way to calculate this so that I can collect the same data for future nights, or am I going to have to guestimate?
Do you know if it's safe to do solar eclipse timelapse with the included solar filter? Europe will have visible eclipse next year.
yes, it is absolutely safe. You can point it at the sun without an eclipse, and an eclipse just has less light :)
@the_space_koala an eclipse normally take a few hours, that's why I'm worried. I've only tried single shots both with my camera and the seestar. I have 15 stop nd filter for the camera.
You really don't need to worry. As long as you use a front-mounted filter made for solar use your equipment will be fine.
Hi whilst this is an exellent tutorial i have found that this teqnique creates ugly banding on my images is thare any reason for this?
I have poseted a more detailed explanation of this on the Pixinsight forum and referenced Space Koala thanks in advance
found it!
What software are you using???
PixInsight
👍
Which one is better for deep sky photography seestar s50 or s30?
It’s a bit like “how long is a piece of string?” My personal choice between these 2 would be the S50 because in general the larger aperture the better
It is nice what is possible with 7h. I wait until my Wife say yes. After that i look forward to 10-20h images to take.
Tell her Christmas is coming 😁 hope you get it soon!
After seeing this process, I will avoid mosaics, LOL
Haha it’s definitely a commitment - otherwise just stick to the single night stack of the seestar 😁
What application did you use for debaying?
Pixinsight