20 replies to this topic

[SkyHunter1]

Hello All,

I will be sending my scope to a remote hosting facility very shortly. I will not be using a rotator and will do mosaics. I have never done mosaics before. I plan to set the chip at 90 degrees and just do mosaics when i need more FOV for a target. I will be using a spacecat 51 and a 2600mc with an AM5 and ASIAIR. I will be using astro pixel processor for stacking, I will also be using the ASIAIR mosacing tool to plan and create these mosaics. How difficult is mosaicing really? it seems daunting to get a 4 panel mosaic for instance to have no seams and be evenly illuminated.

 

I would like to know the finer points of mosaics. For instance, I was told I should take all frames each night and do this every subsequent night. It was explained that this was to ensure that frame 1 would have the same illumination differences as frame 2,3 or 4 for example. i could add more frames each night, but each nights set should be uniformly illuminated when i stack them if each set of 4 subs were taken each night. This seems like sound advice. Another tip was to overlap at least 20 percent. 

 

What do you do if you lose many frames from a single panel (clouds etc) i would imagine that if one frame had less subs that would be an issue to ensure no seams due to brightness differences. I would imagine noise would be different from the other frames. How do you recover from that?

 

Any advice and tips you can give me would be appreciated. I have never done mosaics, so im pretty much a blank slate. If anyone has tips on stacking in APP that would be great as well. 

 

Thanks in advance to this amazing community!!

 

Regards,

Skyhunter1

 

 

[VMan]

I've only stuck my toe in the water with mosaics so can't be much help. But I was surprised how easy it was. Here is my squid/seahorse mosiac: https://flic.kr/p/2qgedSV
The software for making mosaics is sophisticated and does things like normalizing frames. I was using Pi and from what I understand APP is even better and will allow you to make even make mosaics taken with different equipment.

Curious to see what others report

Edited by VMan, 16 February 2025 - 11:44 PM.

[OregonAstronomer]

Hello All,

I will be sending my scope to a remote hosting facility very shortly. I will not be using a rotator and will do mosaics. I have never done mosaics before. I plan to set the chip at 90 degrees and just do mosaics when i need more FOV for a target. I will be using a spacecat 51 and a 2600mc with an AM5 and ASIAIR. I will be using astro pixel processor for stacking, I will also be using the ASIAIR mosacing tool to plan and create these mosaics. How difficult is mosaicing really? it seems daunting to get a 4 panel mosaic for instance to have no seams and be evenly illuminated.

 

I would like to know the finer points of mosaics. For instance, I was told I should take all frames each night and do this every subsequent night. It was explained that this was to ensure that frame 1 would have the same illumination differences as frame 2,3 or 4 for example. i could add more frames each night, but each nights set should be uniformly illuminated when i stack them if each set of 4 subs were taken each night. This seems like sound advice. Another tip was to overlap at least 20 percent. 

 

What do you do if you lose many frames from a single panel (clouds etc) i would imagine that if one frame had less subs that would be an issue to ensure no seams due to brightness differences. I would imagine noise would be different from the other frames. How do you recover from that?

 

Any advice and tips you can give me would be appreciated. I have never done mosaics, so im pretty much a blank slate. If anyone has tips on stacking in APP that would be great as well. 

 

Thanks in advance to this amazing community!!

 

Regards,

Skyhunter1

Hi,

 

I don't know if there is "right way" or not, but I'm happy to let you know what I've done personally with remotely-obtained mosaics over the last couple of years.

 

I also use AstroPixel Processor - its mosaic capabilities are second to none. I use NINA for planning the mosaics, but the ASIAir works similarly.

 

I overlap my panels 40%. That may be a bit excessive for some, but it allows me to get rid of slightly misshapen stars or gradients at the edges.

 

I obtain a full night's data on one panel, then move on to the next panel the next night. I don't like wasting imaging time slewing, centering and plate-solving between panels. I set things in motion on one panel at the beginning of the night and just let 'er rip all night long. After reviewing the subs, if I have to delete any, I go back to that panel later and repeat it. I've never had a problem with the panels not appearing the same or any other artifacts from having obtained data for different panels on different nights.

 

I've done 9-panel mosaics in APP this way with no seams, no uneven illumination nor any other problems (save for the processing time of 900 light, dark, flat and flat-dark frames!).

 

YMMV,

 

Arnie

[SkyHunter1]

 

 

I obtain a full night's data on one panel, then move on to the next panel the next night. I don't like wasting imaging time slewing, centering and plate-solving between panels. I set things in motion on one panel at the beginning of the night and just let 'er rip all night long. After reviewing the subs, if I have to delete any, I go back to that panel later and repeat it. I've never had a problem with the panels not appearing the same or any other artifacts from having obtained data for different panels on different nights.

 

 

Thanks so much Arnie, in refrence to the above, if you lose subs on a particular panel, when you say you go back and repeat it. I assume you repeat the number of lost frames? if so is the goal to have the same number of frames in each panel, so lets say the goal is150 frames per panel. you would redo 20 frames on one panel if you deleted 20 frames? 

 

Also you mention that you havent had any problems with panels not appearing the same. If i do panel one on new moon and i do another panel where the moon is out (however slightly) wont that change the illumination of the panels? I would assume that the conditions would have to be the same for all of the panels for them to match. This is why i thought that the advice to take all panels every night and do this on every subsequent night seemed correct. this way the stack of each panel were stacking the same set of conditons over multiple nights

 

I would much prefer to do a panel a night as you suggest and youve been ok as you mention, but I imagine that conditions would have to be similar for each night. is this accurate?

 

Thanks so much again for your help.

 

Regards,

Skyhunter1

[dswtan]

What I do, FWIW:
 
- Planning and sequencing in NINA. Stacking in PI (WBPP). Batch convert XISFs from WBPP to FIT via ImageContainer. Mosaic alignment in APP. Final processing in PI (back to XISF). Annoying, but it works. And APP is just so good for this. (2.0 Beta 33 just fixed a bunch of useful stuff too.)
 
- 20% frame overlap, per NINA default. Sometimes I might sneak to 15% if I need just a little more room (my sensors have tended to be smaller than APS-C). I try to avoid 10% but may have done it.
 
- I don't worry about misshapen edge stars; BXT takes care of that, especially at 20%. A 10% overlap is probably fine these days, thanks to BXT, but I'll stick to what works for me.
 
- I try to keep panels balanced in exposure by keeping an eye on numbers of subs and general sky conditions, but I don't get paranoid about it at capture. I have 2-3hr capture windows (due to trees, etc.), so while I try to increment all panels the same night, often it's not possible -- especially for more than 2-panels. I'm relying on averaging conditions over multiple nights anyway. So as long as I have a roughly equal number of subs per filter, I'm fine. My local weather is not so variable as to be a big problem, and I avoid imaging if the seeing is bad or other conditions are marginal. 
 
- I manually Blink subs after every session fairly aggressively, so I know my subs are all of roughly decent quality before I stack. Averaging in stacking takes care of any remaining outliers and smooths conditions out. Again, I do make sure I'm stacking roughly equal number of quality subs per panel and filter -- say, <5-10% max of each other in terms of #subs. I've not done any math or experiment to verify this is necessary, just gut feel.
 
- I'm happy with this approach but I don't have too many examples done recently -- much data is in my processing backlog . I tend to take smaller 2- and 4-panel mosaics, not those giant ones. Couple here though, 2-panel last year and 4-panel two years ago:



- And this 2-panel draft of M33 that needs more work before Astrobin: https://zoomhub.net/w29gJ

[SkyHunter1]

Thanks for the advice dswtan! I appreciate the input. Great images by the way

 

Regards,

Skyhunter1

[BucketDave]

I've been doing mosaics this autumn, using SharpCap and APP with my 65phq/ASI533mc. I started on the Bubble+Lobster and moved on to the Heart.

In SharpCap, I've written some sequences that will do 21 minutes per panel (four times 5min exposures), then jump to the next frame. Once all five panels are done, it does an auto-focus and repeats. It gives a gradual build-up of the whole mosaic - that's useful if you want to adjust your framing.

Both targets were five-panel mosaics. Four outers with maybe 10% overlap plus one central panel means that I can use non-mosaic registration in APP which makes that process MUCH faster. It works because every outer panel overlaps the central panel. Just makes sure that the reference frame for registration is one of the images of the central panel. I then integrate all five panels in one operation.

If you do bigger mosaics, you'll need to do mosaic registation and pay the time penalty.

But mosaics are slow. It's making me think of moving from a 533 camera to a 2600 clone.

 

This is my 5-panel Heart in Ha, heavily reduced to get down below 500kB

 

 

Edited by BucketDave, 18 February 2025 - 03:12 PM.

[Exquisitus]

I dipped my toes into mosaics.  Here is my favorite so far.  I have a tiny scope and even tinier camera so was required to do a mosaic to shoot the rosette ****.  Ill also include a 4 frame mosaic of California nebula. But I thought it came out quite nicely and extremely painless in pixinsight.  To me I cant tell its multiple photos and I watched a youtube video and walah. My rosette mosaic was 14 subs on top and 29 on bottom.  Could I have lost data on top?  For sure, but I still felt it was worthwhile. 

Edited by Exquisitus, 18 February 2025 - 06:00 PM.

[Joe G]

Here is a 20 panel mosaic processed with APP. Obviously done over several nights.

 

https://www.flickr.c...157694685889534

[Joe G]

More to your question.  I just follow APP's suggestions for mosaics.  Under 4 in the Register tab you choose "Mosaic."  Then in 6, the Integration tab, I set Multi Band Blending to 20.  When you press the "Integrate" tab APP then will usually ask you if you want to change some setting.  So do that.

 

Start with a simple 2 panel mosaic.  Narrow band is easier.  Then try RGB panels.

[OregonAstronomer]

Thanks so much Arnie, in refrence to the above, if you lose subs on a particular panel, when you say you go back and repeat it. I assume you repeat the number of lost frames? if so is the goal to have the same number of frames in each panel, so lets say the goal is150 frames per panel. you would redo 20 frames on one panel if you deleted 20 frames? 

 

Also you mention that you havent had any problems with panels not appearing the same. If i do panel one on new moon and i do another panel where the moon is out (however slightly) wont that change the illumination of the panels? I would assume that the conditions would have to be the same for all of the panels for them to match. This is why i thought that the advice to take all panels every night and do this on every subsequent night seemed correct. this way the stack of each panel were stacking the same set of conditons over multiple nights

 

I would much prefer to do a panel a night as you suggest and youve been ok as you mention, but I imagine that conditions would have to be similar for each night. is this accurate?

 

Thanks so much again for your help.

 

Regards,

Skyhunter1

Correct, I just repeat the number of lost subs. I like to have the number of subs for each panel be as close as possible, although a few plus or minus in either direction does not seem to matter.

 

I personally only do broadband when the moon is not up, so there is no problem with some panels being affected by moonlight while others are not. When narrowband imaging, as long as I'm not pointing too close to the moon, it doesn't matter!

[ntph]

What I do, FWIW:
 
- Planning and sequencing in NINA. Stacking in PI (WBPP). Batch convert XISFs from WBPP to FIT via ImageContainer. Mosaic alignment in APP. Final processing in PI (back to XISF). Annoying, but it works. And APP is just so good for this. (2.0 Beta 33 just fixed a bunch of useful stuff too.)
 
- 20% frame overlap, per NINA default. Sometimes I might sneak to 15% if I need just a little more room (my sensors have tended to be smaller than APS-C). I try to avoid 10% but may have done it.
 
- I don't worry about misshapen edge stars; BXT takes care of that, especially at 20%. A 10% overlap is probably fine these days, thanks to BXT, but I'll stick to what works for me.
 
- I try to keep panels balanced in exposure by keeping an eye on numbers of subs and general sky conditions, but I don't get paranoid about it at capture. I have 2-3hr capture windows (due to trees, etc.), so while I try to increment all panels the same night, often it's not possible -- especially for more than 2-panels. I'm relying on averaging conditions over multiple nights anyway. So as long as I have a roughly equal number of subs per filter, I'm fine. My local weather is not so variable as to be a big problem, and I avoid imaging if the seeing is bad or other conditions are marginal. 
 
- I manually Blink subs after every session fairly aggressively, so I know my subs are all of roughly decent quality before I stack. Averaging in stacking takes care of any remaining outliers and smooths conditions out. Again, I do make sure I'm stacking roughly equal number of quality subs per panel and filter -- say, <5-10% max of each other in terms of #subs. I've not done any math or experiment to verify this is necessary, just gut feel.
 

Pretty much the same approach, except I find I am quite happy using PI's PMM scripts and workflow. Maybe APP is simpler and faster, but I would be very surprised if it gave better results. My first attempt was with GradientMergeMosaic--nasty artefacts. Same data with PMM--flawless. I have never looked back. Except after shooting quite a few mosaics, I am much happier to shoot widefield images now rather than shoot mosaics! Having said that, now I am planning widefield mosaics, even with using a 100-400 mm lens. Can't win. 

[astromiester1]

Has anyone tried creating a mosaic with 2 different perm mounts mounted side by side with identical optics instead of 2 OTAs on the same mount?

 

I have 2 Software Bisque mounts with identical optical trains within 3 feet of each other and want to create a simple mosaic image using both to create a larger feild image on a target in question.

They both have hyperstars mounted on SB mounts with C11 OTAs and ZWO 2600MC cameras for instance. I should be able to do this but most of the research I see assumes you have 2 OTAs on one mount. I would think the process is the same perhaps using pixinsight but was curious if anyone has attempted this to cut in half the time to create a Mosaic using only 1 mount.

 

Each mount by the way has its own computer running SKYX.

 

Thanks and Clear Skys...

Edited by astromiester1, 22 February 2025 - 09:04 PM.

[psandelle]

As long as you're plate-solving both mounts to identical spots in the mosaic, then it should be no problem at all. If I were using something like Voyager, I'd have one computer running both rigs and that would link them both up just fine. I'm sure NINA could do that on two mounts, I believe (but check, as I've only seen it done with two scopes on one mount).

Paul

[ntph]

No idea about running two separate imaging trains from one computer, if that's what you're asking. Otherwise, assuming  that  each of your rigs has its own computer driving the bus it would be very straightforward. Create your mosaic instructions  and have one rig shoot just one panel and the other rig shoos the other panel. Treat all the calibrated data as "normal" mosaic set. I don't see any issues with this. Three feet of parallax...

[psandelle]

No idea about running two separate imaging trains from one computer, if that's what you're asking. Otherwise, assuming  that  each of your rigs has its own computer driving the bus it would be very straightforward. Create your mosaic instructions  and have one rig shoot just one panel and the other rig shoos the other panel. Treat all the calibrated data as "normal" mosaic set. I don't see any issues with this. Three feet of parallax...

I'd probably go with both scopes shooting the same mosaic panels at the same time. That's sort of a no-brainer and takes less software thought. You can do it the other way, but it takes a couple more steps (well, at least one with Voyager, not sure about NINA or Prism).

Paul

[astromiester1]

Hi Paul, thats a great idea... So, 2x on each panel then both move to next panel and so on then combine them all... 

 

Thanks... I will try and figure all this out...

 

Dale

[dswtan]

Pretty much the same approach, except I find I am quite happy using PI's PMM scripts and workflow. Maybe APP is simpler and faster [...]

Also, APP is only really an easy choice for me since I already paid for it and have some basic familiarity with its own quirky interface. I would hesitate to purchase it only for mosaics. That said, it is *that* good and easy, in my personal experience. 

 

I re-looked at PI's options. Oh, I see PMM is a third party and requires NSG? That's already way too much changing workflow for me. But thanks for the prompt.

 

Ref: https://astroprocessing.com/ [PMM, NSG]
 

[ntph]

Also, APP is only really an easy choice for me since I already paid for it and have some basic familiarity with its own quirky interface. I would hesitate to purchase it only for mosaics. That said, it is *that* good and easy, in my personal experience. 

 

I re-looked at PI's options. Oh, I see PMM is a third party and requires NSG? That's already way too much changing workflow for me. But thanks for the prompt.

 

Ref: https://astroprocessing.com/ [PMM, NSG]
 

PMM doesn't require NSG; they are both included in the repository and will be loaded as scripts. NSG will run at a certain level as a free script. PMM is free regardless of whether you upgrade with a license of NSGXnml. 

 

From what I have read about APP, I suspect the workflow  shift to PMM would be much different and perhaps too much so for some. But it really isn't hard to use at all, especially if one follows the simple and quite explicit instructions provided. And if you run into difficulty with anything, the extensive documentation provided will get you through just about anything. And should that fail, the author John Murphy is frequently on the PI Forum and answers questions there. Once you understand the workflow, it is really very simple and fast. For most data sets, the default settings work just fine.   

 

I feel a bit sheepish as a booster for PMM since I don't have any experience with other non-PI options. But I haven't found a head to head comparison anywhere, either. All I know is that for my first attempt, I used PI's Gradient Merge Mosaic and it was full of pinch artefacts and seams. I couldn't get it cleaned up very well. Somehow I found out about the PMM script and I ran the same data through it at defaults and the result was essentially perfect. Subsequent use on other projects taught me a few things, all solved by looking at the documentation. Using PMM you can create mosaics in mere seconds, essentially as quickly as you can tell each script which images you want to use. The rate limiting step in PI is having an astrometric solution for each panel. Once you have that, it is a simple matter of telling first MosaicByCoordinates which panels you are using, then  use TrimMosaicTile to do its thing to each of the resulting tiles from MosaicByCoodinates, then using PMM to join these trimmed tiles. A 2 panel mosaic is the simplest. 3 would be next: join panel 1 and panel 2, then join panel 3 to the new image just created (which is called Mosaic). To do a 2x2, you would join panels that share a common side in pairs then join those two new panels. Sounds more complicated than it is. Maybe you can find the time and inclination to try it out and let us know your impressions. I for one would be interested in a real user comparison. 

[astromiester1]

Thanks you guys for all chimming in with great information...I think I will stick with PI for the processing..Can't wait to try this out...

 

Dale

Edited by astromiester1, 24 February 2025 - 07:43 AM.

[CosmicWreckingBall]

Once upon a time I created a 200 panel mosaic over the course of 7 years.  

 

https:www.spaceforeverybody.com/Orion 

 

The coded interface tool in the top left allows you to see the RGB, HA, and Plate Solved data individually by clicking the color coded buttons. 

 

So I have some serious PTSD when it comes to the “M” word.  Regardless, I used just about every process aligning and processing my data to find the best.  In my case with so much data, once I used photomerge mosaic and mosaic by coordinates , all else was inferior. 

 

Since then, I will say Astro Pixel Processor has a really good mosaic tool that’s intuitive and great for 4-8 panel mosaics. I’ve seen larger mosaics completed with it, but I’ve been hesitant to do larger mosaics again. Ha. 
 

Good luck!

Edited by CosmicWreckingBall, 24 February 2025 - 05:37 PM.