'5 Dice Method'! Making a mosaic for (perfect) round stars and better star color. - Forums

Hi,

since I restarted this beautiful hobby a few years ago, I struggled a lot with different telescopes. I even returned one that had too much chromatic aberration. Like almost everyone, I have spent most of the time trying to find the right backfocus, check collimation, fix guiding errors (differential flexure), went with an OAG, tilt etc. But still,... poor stars shapes and color in all corners. In the end, I got to the point that I just couldn't improve and I just had to live with it.

But, since it got so much easier to make mosaics (Asiair, Telescopius etc.), I thought about this,... It only has to do with the stars, which only need some short exposures. What if I make pictures of the corners separately, as a mosaic and then combine those, how will the star shape be then? Of course, you will have to crop this mosaic and align stars etc. It takes some more time to capture and process, but in the end... See for yourself ;-) I am probably just doing what many other already did and know for many years but I havent thought of it before. I would call this the '5 Dice Method'.

This is the first test I did with a Sharpstar 94edph and ASI2600MM-pro

Before:

After:

Final image:
M45, Pleiades

Arnaldo

The "after" image lacks faint stars. And when you go deep for nebulas, you will end up with round bright stars, and still lousy faint stars. If you don't go deep, then this can be a solution.

This is actually a clever idea. You could register the stars mosaic with the original field of view to ensure that the stars are exactly where they belong. Then remove all stars from the original and just add in the mosaic stars after. I like it. Of course the DSO will still be distorted but it wouldn't stand out like an aberrated point source.

This is actually a clever idea. You could register the stars mosaic with the original field of view to ensure that the stars are exactly where they belong. Then remove all stars from the original and just add in the mosaic stars after. I like it. Of course the DSO will still be distorted but it wouldn't stand out like an aberrated point source.

Exactly!

Wei-Hao Wang:
The "after" image lacks faint stars. And when you go deep for nebulas, you will end up with round bright stars, and still lousy faint stars. If you don't go deep, then this can be a solution.

I have to see how it works out on other objects but for now I’m happy with the amount of stars that I catch within 30 seconds. You can of course image longer if you want and also make a long exposure mosaic, including the nebula). That would only take you almost 5x as long 🥴 I always end up making the stars less prominent in post processing and maybe loose very faint stars anyway…

I think it will end up with the ultimate question of how much time you have. I have limited imaging time in dark places, so I would rather prefer saving some money and get a good scope that covers a big sensor.

Then my ultimate question would be,… how much money do you have 🤣🥴 I came up with this solution because I wanted to get the best out of what I have. And what I have here is what most amateurs would have… Not a Takahashi but Sharpstar or TS apo. And a modest Skywatcher or iOptron mount, Zwo camera etc… Expensive enough for most of us amateurs. Not to compare with $50.000+ equipment if you ask me…

I dont think the exercise of collecting the stars mosaic would be that time consuming. In an hour you could cover all 4 corners and that would be enough.

Wei-hao is correct that you won't record faint stars. It really depends on your goals. I'm finding that in my recent processing I'm going deep for the nebula but pull stars out while at the linear stage and process separately. So even in my 10 to 20 hour integrations with epsilon the faint stars are not visible in my image even though I captured them.

It's really all about your goals. I personally am just here to enjoy the hobby and make pretty pictures. If I was trying to see what mag stars I could bring out that would be a different story.

Your idea is a good one and for those people who bought an imx571 or inx455 chip but dont have a scope that can correct the full field this would be an elegant solution.

I think it is important though that you disclose in your final image that you mosaiced the stars. A lot of people use astrobin as a resource to evaluate equipment so if you show perfect stars to the corners of a budget scope it could be misleading.

Wei-Hao Wang:
I have limited imaging time in dark places, so I would rather prefer saving some money and get a good scope that covers a big sensor.

Even in this case, you have to worry about optimizing your entire imaging train (tilt, backspacing etc.) to get perfect stars at the corners, so an expensive scope is not always a perfect solution.

Hi Arnaldo,

For its a very clever idea. I don't know whether many people "already did this". I personally didn't think about it. Your before and after example proves your concept will help.
Of course Wei-Hao us right. You lost many smaller stars. But I guess this challenge can be fixed by optimizing exposure and stretching of the "star image".
Thanks for sharing
Thomas

Arun H:
Wei-Hao Wang:
I have limited imaging time in dark places, so I would rather prefer saving some money and get a good scope that covers a big sensor.

Even in this case, you have to worry about optimizing your entire imaging train (tilt, backspacing etc.) to get perfect stars at the corners, so an expensive scope is not always a perfect solution.

Actually, I never spent time on tilt or backspacing problems. All my scopes and cameras are spot on in terms of that. On the other hand, when I use a reflector, I do spend time on collimation, but the time on collimation do not consume my imaging time.

I think here the point is that I start with a scope that I know it will produce good stars in the corners. Of course, Arnaldo made a point about the lack of budget. I think that's a fair point.

Thomas:
Hi Arnaldo,

For its a very clever idea. I don't know whether many people "already did this". I personally didn't think about it. Your before and after example proves your concept will help.
Of course Wei-Hao us right. You lost many smaller stars. But I guess this challenge can be fixed by optimizing exposure and stretching of the "star image".
Thanks for sharing
Thomas

I was not after the ‘faint stars’ but that wouldn’t be a problem to fix I think. ;-)

I dont think the exercise of collecting the stars mosaic would be that time consuming. In an hour you could cover all 4 corners and that would be enough.

Wei-hao is correct that you won't record faint stars. It really depends on your goals. I'm finding that in my recent processing I'm going deep for the nebula but pull stars out while at the linear stage and process separately. So even in my 10 to 20 hour integrations with epsilon the faint stars are not visible in my image even though I captured them.

It's really all about your goals. I personally am just here to enjoy the hobby and make pretty pictures. If I was trying to see what mag stars I could bring out that would be a different story.

Your idea is a good one and for those people who bought an imx571 or inx455 chip but dont have a scope that can correct the full field this would be an elegant solution.

I think it is important though that you disclose in your final image that you mosaiced the stars. A lot of people use astrobin as a resource to evaluate equipment so if you show perfect stars to the corners of a budget scope it could be misleading.

I use Astrobin as a source for comparison also and have always contacted the photographer whenever I wondered why their stars were better or their pictures were sharper. In those discussions I have learned a lot. I learned about differential flexure and went with an OAG from that point (with my Edge Hd8). Also the use of Topaz software isn’t a secret or forbidden anymore. I might mention my method though, as I like to share all secrets I have learned so far ;-).

Wei-Hao Wang:
Arun H:
Wei-Hao Wang:
I have limited imaging time in dark places, so I would rather prefer saving some money and get a good scope that covers a big sensor.

Even in this case, you have to worry about optimizing your entire imaging train (tilt, backspacing etc.) to get perfect stars at the corners, so an expensive scope is not always a perfect solution.

Actually, I never spent time on tilt or backspacing problems. All my scopes and cameras are spot on in terms of that. On the other hand, when I use a reflector, I do spend time on collimation, but the time on collimation do not consume my imaging time.

I think here the point is that I start with a scope that I know it will produce good stars in the corners. Of course, Arnaldo made a point about the lack of budget. I think that's a fair point.

I have a lack of budget and just don’t like to spend too much time adjusting backfocus or what so ever for just very little improvement. So, that’s why I came up with this idea. Another option I might want to try is combining stars from my Sharpstar with nebulas from my Edge HD8. 414mm vs 1460mm. Another approach, even quicker and easier.

I dunno, this is a bandaid to hide a fundamental problem. You've got to be careful to fit the stars so that they fit the outer part of the field with regard to optical distortion. That lousy looking PSF also applies to whatever object you are imaging as well. If you don't care about sharp imaging, then simply replacing the distorted stars with round stars works, but in my view, it's not a very good solution. I'm in the camp that would sell the bad scope and buy a better one to fix this sort of problem.

John

John Hayes:
I dunno, this is a bandaid to hide a fundamental problem. You've got to be careful to fit the stars so that they fit the outer part of the field with regard to optical distortion. That lousy looking PSF also applies to whatever object you are imaging as well. If you don't care about sharp imaging, then simply replacing the distorted stars with round stars works, but in my view, it's not a very good solution. I'm in the camp that would sell the bad scope and buy a better one to fix this sort of problem.

John

I already did that,… returned a (really) bad scope and got this one. What is ‘bad’…? And is it a ‘problem’? I think this scope is a great scope for the price. I just don’t have the budget for a Takahashi… and find this a great option to fix at least one thing I sometimes don’t like about my images. The loss of sharpness in the object itself is so much less noticeable, I don’t bother about that. But, I could fix that also by just making a mosaic with very big overlaps all in all. I agree that buying a better scope is the best option but that was and is just not an option. I look at it as a musician. I am a guitarist, but never bought very expensive guitars. That doesn’t mean that my tracks sound bad, I just had to be more creative and do my very best to make it sound like an expensive guitar. If I had the money I would definitely buy a very good, expensive guitar, don’t get me wrong. Although,… I am able to build them myself now 😎😜 Not that I will build a telescope now… although 🤔🤣

John Hayes:
I dunno, this is a bandaid to hide a fundamental problem. You've got to be careful to fit the stars so that they fit the outer part of the field with regard to optical distortion. That lousy looking PSF also applies to whatever object you are imaging as well. If you don't care about sharp imaging, then simply replacing the distorted stars with round stars works, but in my view, it's not a very good solution. I'm in the camp that would sell the bad scope and buy a better one to fix this sort of problem.

John

Hello John,

I struggle with this issue of star reduction. I use Starnet to extract them but it sometimes leaves artifacts and then reduce the number of stars and put them back in but I'm never really happy with the results. I've also tried the photoshop method of creating a mask etc. Is there a better way to do this? or do I just need to get better with these methods. Any ideas would be appreciated. Thanks

If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic… In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

Yuxuan:
If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic... In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

I know that buying a more expensive scope could fix this issue, it ‘could’. But that’s just the whole point. I can’t afford such telescopes! The most simple solution would be to just crop the image or image with a smaller sensor. This mosaic method was just an idea to get the most and maybe even more out of what I have. I am not concerned about the nebula’s sharpness in the corners. I just don’t like the bird shaped stars that much… And because it is just the stars that bother me, it doesn’t take me that long for the mosaic. Half an hour or maybe an hour per frame would be enough. If someone could tell me which telescope around 500mm focal length will give me perfect round stars with a crop sensor under €1500… I would be happy to know about it!

Arnaldo Lopez:
Yuxuan:
If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic... In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

I know that buying a more expensive scope could fix this issue, it ‘could’. But that’s just the whole point. I can’t afford such telescopes! The most simple solution would be to just crop the image or image with a smaller sensor. This mosaic method was just an idea to get the most and maybe even more out of what I have. I am not concerned about the nebula’s sharpness in the corners. I just don’t like the bird shaped stars that much… And because it is just the stars that bother me, it doesn’t take me that long for the mosaic. Half an hour or maybe an hour per frame would be enough. If someone could tell me which telescope around 500mm focal length will give me perfect round stars with a crop sensor under €1500… I would be happy to know about it!

As you know, I think your method is sound... but to answer your last question the AT92/TMB92 with an Esprit 100 Field Flattener will produce round stars to the corners of APS-C. It's 500mm FL. Cost is a little higher than 1500 but not much, and they can be had on the used market for less. Optically they are excellent and I would not hesitate to pair it with IMX571.

Arnaldo Lopez:
Yuxuan:
If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic... In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

I know that buying a more expensive scope could fix this issue, it ‘could’. But that’s just the whole point. I can’t afford such telescopes! The most simple solution would be to just crop the image or image with a smaller sensor. This mosaic method was just an idea to get the most and maybe even more out of what I have. I am not concerned about the nebula’s sharpness in the corners. I just don’t like the bird shaped stars that much… And because it is just the stars that bother me, it doesn’t take me that long for the mosaic. Half an hour or maybe an hour per frame would be enough. If someone could tell me which telescope around 500mm focal length will give me perfect round stars with a crop sensor under €1500… I would be happy to know about it!

As you know, I think your method is sound... but to answer your last question the AT92/TMB92 with an Esprit 100 Field Flattener will produce round stars to the corners of APS-C. It's 500mm FL. Cost is a little higher than 1500 but not much, and they can be had on the used market for less. Optically they are excellent and I would not hesitate to pair it with IMX571.

Thanks for the tip! Who knows…someday… The telescopes I own right now will probably not be the last ones 🥴🤑

Arnaldo Lopez:
Arnaldo Lopez:
Yuxuan:
If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic... In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

I know that buying a more expensive scope could fix this issue, it ‘could’. But that’s just the whole point. I can’t afford such telescopes! The most simple solution would be to just crop the image or image with a smaller sensor. This mosaic method was just an idea to get the most and maybe even more out of what I have. I am not concerned about the nebula’s sharpness in the corners. I just don’t like the bird shaped stars that much… And because it is just the stars that bother me, it doesn’t take me that long for the mosaic. Half an hour or maybe an hour per frame would be enough. If someone could tell me which telescope around 500mm focal length will give me perfect round stars with a crop sensor under €1500… I would be happy to know about it!

As you know, I think your method is sound... but to answer your last question the AT92/TMB92 with an Esprit 100 Field Flattener will produce round stars to the corners of APS-C. It's 500mm FL. Cost is a little higher than 1500 but not much, and they can be had on the used market for less. Optically they are excellent and I would not hesitate to pair it with IMX571.

Thanks for the tip! Who knows…someday… The telescopes I own right now will probably not be the last ones 🥴🤑

EDIT: Never mind I misunderstood your current situation. Cheers!
I don't know for sure but it seems like you C-Edge should be able to produce round stars. Are you sure it's collimated and all that jazz? Maybe my expectations are out of line.

I don't know for sure but it seems like you C-Edge should be able to produce round stars. Are you sure it's collimated and all that jazz? Maybe my expectations are out of line.

500mm focal length was the criteria.

Arnaldo Lopez:
Arnaldo Lopez:
Yuxuan:
If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic... In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

I know that buying a more expensive scope could fix this issue, it ‘could’. But that’s just the whole point. I can’t afford such telescopes! The most simple solution would be to just crop the image or image with a smaller sensor. This mosaic method was just an idea to get the most and maybe even more out of what I have. I am not concerned about the nebula’s sharpness in the corners. I just don’t like the bird shaped stars that much… And because it is just the stars that bother me, it doesn’t take me that long for the mosaic. Half an hour or maybe an hour per frame would be enough. If someone could tell me which telescope around 500mm focal length will give me perfect round stars with a crop sensor under €1500… I would be happy to know about it!

As you know, I think your method is sound... but to answer your last question the AT92/TMB92 with an Esprit 100 Field Flattener will produce round stars to the corners of APS-C. It's 500mm FL. Cost is a little higher than 1500 but not much, and they can be had on the used market for less. Optically they are excellent and I would not hesitate to pair it with IMX571.

Thanks for the tip! Who knows…someday… The telescopes I own right now will probably not be the last ones 🥴🤑

I don't know for sure but it seems like you C-Edge should be able to produce round stars. Are you sure it's collimated and all that jazz?

The Edge hd is fine, this is all about the Sharpstar 94EDPH. Which is a nice scope also, don’t get me wrong. And probably could make the stars better if I fine-tune it a little more. I remember that I used to have better star shapes before… I have to admit that I lean towards the Edge hd8 now for imaging. I also have a C11 (classic) which I am going to use for that as soon as I get the Starizona LF reducer. I tend to like the longer focal lengths a bit more…

Arnaldo Lopez:
Arnaldo Lopez:
Yuxuan:
If you only fix the stars, the fundamental problem, which is the blurriness of the deep sky objects you are trying to shoot is still there. To me the bad corner stars is only a symptom used to diagnose a fundamental problem.

I like the idea though, if only it doesn't take a goddarn long time to make a mosaic... In principle one could completely solve the optical problem by taking many shots of overlapping regions, crop out the peripherals, and then sew together the centers of each image. However, it would take 9 times the integration and processing time (if one only keeps the center 1/3 in each direction), not counting the tremendous amount of extra time for mosaic. Plus if any of the 9 frames were bad, the whole mosaic would be ruined.

I promise you if you do this a few times, you will find a way to afford a more expensive scope.

I know that buying a more expensive scope could fix this issue, it ‘could’. But that’s just the whole point. I can’t afford such telescopes! The most simple solution would be to just crop the image or image with a smaller sensor. This mosaic method was just an idea to get the most and maybe even more out of what I have. I am not concerned about the nebula’s sharpness in the corners. I just don’t like the bird shaped stars that much… And because it is just the stars that bother me, it doesn’t take me that long for the mosaic. Half an hour or maybe an hour per frame would be enough. If someone could tell me which telescope around 500mm focal length will give me perfect round stars with a crop sensor under €1500… I would be happy to know about it!

As you know, I think your method is sound... but to answer your last question the AT92/TMB92 with an Esprit 100 Field Flattener will produce round stars to the corners of APS-C. It's 500mm FL. Cost is a little higher than 1500 but not much, and they can be had on the used market for less. Optically they are excellent and I would not hesitate to pair it with IMX571.

Thanks for the tip! Who knows…someday… The telescopes I own right now will probably not be the last ones 🥴🤑

Also consider the Esprit 80 and the RASA 8. Both have 400mm of FL, are ~$2000, and should handle APS-C with no issues. For the RASA you would need e.g., a Baader tilt adjuster to ensure sensor orthogonality.