M13 RGB

[vlaiv]

1 minute ago, Rodd said:

The bright star is larger in binned one.  And the resolution is not as high.  Looking at the edges between light and dark regions you can see they are less sharp in the binned.   But this just might mean 4x4 was too much.  3x3 maybe would be perfect.

I did this example on 8bit stretched data, so maybe there will be very small difference - but in reality, what you see as noise can be perceived as sharpness - but it is really noise. Mentally we perceive a bit of noise as sharpness - because blurriness looses that "edge" of individual pixels.

2 minutes ago, Rodd said:

But this aside--if you use the software binning on board the ASI 1600, you actually get 10 bit readout.  Some say it is better to collect 1x1 and software bin after on the computer.  Would you agree?  They also say its ok to software bin after noise reduction and some processing.  I thought binning should be done in linear state.  

I'm for software binning after acquisition. Best time to bin is after calibration and before integration. Best way to bin is "split" bin.

Simplest way to do it is to just integrate as is and then bin result. In fact - that is what you can do if you have stacked linear data - just bin that.

In any case - binning should be done in linear state. No noise reduction / no processing.

[Rodd]

37 minutes ago, vlaiv said:

I did this example on 8bit stretched data, so maybe there will be very small difference - but in reality, what you see as noise can be perceived as sharpness - but it is really noise. Mentally we perceive a bit of noise as sharpness - because blurriness looses that "edge" of individual pixels.

I'm for software binning after acquisition. Best time to bin is after calibration and before integration. Best way to bin is "split" bin.

Simplest way to do it is to just integrate as is and then bin result. In fact - that is what you can do if you have stacked linear data - just bin that.

In any case - binning should be done in linear state. No noise reduction / no processing.

So the best way to proceed would be to collect at 1x1, calibrate, register, integrate--and bin the individual stacks?  The reason I ask is because I have darks and flats for bin 2 using camera.  If I go to Bin 1 and software bin I will need all new flats and darks.  Is the difference likely to be significant between using camera to bin and doing it myself after collecting 1x1?

Edited May 17, 2021 by Rodd

[vlaiv]

1 hour ago, Rodd said:

So the best way to proceed would be to collect at 1x1, calibrate, register, integrate--and bin the individual stacks?  The reason I ask is because I have darks and flats for bin 2 using camera.  If I go to Bin 1 and software bin I will need all new flats and darks.  Is the difference likely to be significant between using camera to bin and doing it myself after collecting 1x1?

If you already have bin x2 data - then keep it and bin x2 additionally in software.

[Rodd]

5 minutes ago, vlaiv said:

If you already have bin x2 data - then keep it and bin x2 additionally in software.

You misunderstand - I will be starting a new project, so I will be collecting all new data.  Will there be a huge difference between binning on camera and binning on computer.  I ask because I have all the flat and flat darks and darks I will need if I bin on camera.   But if it would be better to bin on computer and collect data using bin 1, then I can bin the integrated stacks--providing that is sufficient and I don't have to bin each sub before integrating.

[vlaiv]

2 minutes ago, Rodd said:

You misunderstand - I will be starting a new project, so I will be collecting all new data.  Will there be a huge difference between binning on camera and binning on computer.  I ask because I have all the flat and flat darks and darks I will need if I bin on camera.   But if it would be better to bin on computer and collect data using bin 1, then I can bin the integrated stacks--providing that is sufficient and I don't have to bin each sub before integrating.

If you are starting a new project, I think it makes more sense to go with bin x1 on integration and later bin to the level it suits you.

You might find that seeing was excellent and you only need to bin x3, or maybe seeing was poor and you bin x5 instead of x4.

[Rodd]

3 minutes ago, vlaiv said:

If you are starting a new project, I think it makes more sense to go with bin x1 on integration and later bin to the level it suits you.

You might find that seeing was excellent and you only need to bin x3, or maybe seeing was poor and you bin x5 instead of x4.

In PI the option to integer resample is 2 up or down.  So to bin 4 do I bin 2 down twice?  No way to do 3.  Its 2 only.  There is a resampling tool but it changes the number of pixels as opposed to integer resample

[vlaiv]

1 minute ago, Rodd said:

In PI the option to integer resample is 2 up or down.  So to bin 4 do I bin 2 down twice?  No way to do 3.  Its 2 only.  There is a resampling tool but it changes the number of pixels as opposed to integer resample

I'm almost certain that you can select 2,3,4,5 ... in integer resample in PI.

Just change outlined factor to wanted bin factor - 2 for x2, 3 for x3, 4 to x4, etc ...

[Rodd]

8 minutes ago, vlaiv said:

I'm almost certain that you can select 2,3,4,5 ... in integer resample in PI.

Just change outlined factor to wanted bin factor - 2 for x2, 3 for x3, 4 to x4, etc ...

Ah...I must have been thinking of 1.5, 2.5 etc.  Thanks.  

[Rodd]

4 hours ago, vlaiv said:

For same total imaging time - less longer subs wins over more shorter subs. This is because of read noise.

Before I forget--not true with 12 bit CMOS cameras, where stacking many subs is the way to go.  

[Rodd]

9 hours ago, vlaiv said:

ust change outlined factor to wanted bin factor - 2 for x2, 3 for x3, 4 to x4, etc ...

Just for kiks I used date from TOA 130 and C11edge--8 hours TOA about 5 hours C11.  

[astrosathya]

Amazing photo

[Rodd]

22 minutes ago, astrosathya said:

Amazing photo

Thanks astrosathya!

[ollypenrice]

Before reading your write-up I was instantly struck by the excellent colour separation in the stars as well as by the fact that they were well resolved. That suggests that your intentions were realized. OK, the stars could be smaller and tighter but that requires the seeing to be at its best. They are hardly shabby as they stand.

When you said that you had variable seeing I found myself wondering what might happen if you made a stack of subs from the best seeing, irrespective of colour, and applied that as luminance, even only partially. Obviously you couldn't do that with a nebula but, on a globular, I wonder if you could get away with it?

Olly

[Mr Spock]

Very impressive images 👍

[geoflewis]

Rodd, I've been following this thread and your several version of this image on Astrobin with great interest. They are all superb images of one of my favourite visual and photographic targets. Your earlier discussions with Vlaiv got me reviewing my most recent image of M13 (https://www.astrobin.com/full/347656/F/), taken with my C14 plus Optec Telecompressor (at ~F7),  which I discovered was taken a few years ago in 2018, so maybe I need to try it again. It got me looking at how well that triplet of gold stars, that Vlaiv discussed, were resolved in my image, which comprised all 120s subs. What I really noticed was how little colour was in the core stars (particularly the blue ones) of my image compared with you original RGB images posted at the start of this thread. My image is much more like your lastest one which included data off the TOA 130. There are clearly more stars revealed in the outer regions of the cluster in your latest combo image, but the core is much brighter, with less colourful and well defined stars, so I was wondering what were the sub exposure lengths with the TOA? Whatever, it is still a cracking image and I hasten to add that it's in a completely different league to mine.

Edited May 18, 2021 by geoflewis

[Rodd]

28 minutes ago, geoflewis said:

Rodd, I've been following this thread and your several version of this image on Astrobin with great interest. They are all superb images of one of my favourite visual and photographic targets. Your earlier discussions with Vlaiv got me reviewing my most recent image of M13 (https://www.astrobin.com/full/347656/F/), taken with my C14 plus Optec Telecompressor (at ~F7),  which I discovered was taken a few years ago in 2018, so maybe I need to try it again. It got me looking at how well that triplet of gold stars, that Vlaiv discussed, were resolved in my image, which comprised all 120s subs. What I really noticed was how little colour was in the core stars (particularly the blue ones) of my image compared with you original RGB images posted at the start of this thread. My image is much more like your lastest one which included data off the TOA 130. There are clearly more stars revealed in the outer regions of the cluster in your latest combo image, but the core is much brighter, with less colourful and well defined stars, so I was wondering what were the sub exposure lengths with the TOA? Whatever, it is still a cracking image and I hasten to add that it's in a completely different league to mine.

Thanks GeoFlewis. They were 5 minute exposures.  For the latest image the stars are much less bloated. I was aiming for realistic.  How does a globular cluster look.  The core is a raging inferno, with nothing resolvable. So I did not process the core as much.  I wanted the core stars visible, but not like they have been painted on. I wanted them hunted at. Also, except for a few of the brightest stars, globular clusters are not very saturated. I think some of my earlier images were too colorful. I probably could have spent more time polishing the image. To tell the truth, I had had enough!  Maybe I am running in circles though. I think the latest image is far superior to the earlier ones due to the stars being smaller, and due to many more stars being revealed throughout.  Maybe it’s all subjective, I don’t know. You like the earliest images?  If you compare my M13 to the greats, which version is closer.  Maybe I need to reassess my judgement.  

[Rodd]

2 hours ago, ollypenrice said:

Before reading your write-up I was instantly struck by the excellent colour separation in the stars as well as by the fact that they were well resolved. That suggests that your intentions were realized. OK, the stars could be smaller and tighter but that requires the seeing to be at its best. They are hardly shabby as they stand.

When you said that you had variable seeing I found myself wondering what might happen if you made a stack of subs from the best seeing, irrespective of colour, and applied that as luminance, even only partially. Obviously you couldn't do that with a nebula but, on a globular, I wonder if you could get away with it?

Olly

Thanks Olly.  You are a mind reader!  I did just that. But it turned out that for the latest image the TOA green stack was by far the sharper then the celestron best subs. Seeing that night must have been Very good. So I used the green stack as a luminance for the last image as I did long ago.  I saw a big change in star tightness once that luminance was added. Maybe I should try it just Using celestron data, though. 

[geoflewis]

26 minutes ago, Rodd said:

Thanks GeoFlewis. They were 5 minute exposures.  For the latest image the stars are much less bloated. I was aiming for realistic.  How does a globular cluster look.  The core is a raging inferno, with nothing resolvable. So I did not process the core as much.  I wanted the core stars visible, but not like they have been painted on. I wanted them hunted at. Also, except for a few of the brightest stars, globular clusters are not very saturated. I think some of my earlier images were too colorful. I probably could have spent more time polishing the image. To tell the truth, I had had enough!  Maybe I am running in circles though. I think the latest image is far superior to the earlier ones due to the stars being smaller, and due to many more stars being revealed throughout.  Maybe it’s all subjective, I don’t know. You like the earliest images?  If you compare my M13 to the greats, which version is closer.  Maybe I need to reassess my judgement.  

Ok so 5 minutes is a bit different to 20s 😏, even with the TOA 130. I agree the latest image is the best of the ones you posted, with much tighter stars, but I did like seeing more colour in the core stars of the 20s RGB versions, of which the 2nd, 'more relaxed' version was my preference. This is a target that I'll spend a long time with visually most years, by long time I mean 30-45 mins sessions just drinking in the view, trying to tease out more stars. I find my TV 12mm gives me the best views, but I now have a binoviewer with a pair of 19mm panoptics, so I'm looking forward to trying those on it. I don't think I've ever seen much in the way of colour at the eyepiece and I find that it is extremely difficult, nigh impossible to resolve stars at the core, though maybe every so often, one or two resolve for a second or two. I think my version is too colourful and recall that I messed with colour star masks and saturation to give them more 'pop', but am now fairly sure that I should have done less, or left them alone completely. Like I said, I've found this thread very interesting and informative, so thanks for that.

Cheers,

Edited May 18, 2021 by geoflewis

[Rodd]

23 minutes ago, geoflewis said:

Ok so 5 minutes is a bit different to 20s 😏, even with the TOA 130. I agree the latest image is the best of the ones you posted, with much tighter stars, but I did like seeing more colour in the core stars of the 20s RGB versions, of which the 2nd, 'more relaxed' version was my preference. This is a target that I'll spend a long time with visually most years, by long time I mean 30-45 mins sessions just drinking in the view, trying to tease out more stars. I find my TV 12mm gives me the best views, but I now have a binoviewer with a pair of 19mm panoptics, so I'm looking forward to trying those on it. I don't think I've ever seen much in the way of colour at the eyepiece and I find that it is extremely difficult, nigh impossible to resolve stars at the core, though maybe every so often, one or two resolve for a second or two. I think my version is too colourful and recall that I messed with colour star masks and saturation to give them more 'pop', but am now fairly sure that I should have done less, or left them alone completely. Like I said, I've found this thread very interesting and informative, so thanks for that.

Cheers,

Once must be very careful with color.  Except for a few unusual examples, space is not a very colorful place.  Processing is a bunch of compromises.  For me, lately, I tend to want as realistic as I can make it.  I equate realistic to what it looks like in an eyepiece.  At least an element of that.  Obviously not completely becuase then our images would be pretty bland.  I touched up the core and tightened up the stars.  This version, I think, is a bit better.  Maybe too much color.  (Most of the added color came from dynamic range modification and not saturation boosts)

[geoflewis]

22 minutes ago, Rodd said:

Once must be very careful with color.  Except for a few unusual examples, space is not a very colorful place.  Processing is a bunch of compromises.  For me, lately, I tend to want as realistic as I can make it.  I equate realistic to what it looks like in an eyepiece.  At least an element of that.  Obviously not completely becuase then our images would be pretty bland.  I touched up the core and tightened up the stars.  This version, I think, is a bit better.  Maybe too much color.  (Most of the added color came from dynamic range modification and not saturation boosts)

Yes, the image gets better and better Rodd. I think your opening comment about colour is completely correct and on point, one must be VERY careful, something that I'm now coming to terms with. Personally I think you have achieved your goal of producing an image that looks like what I see at the eyepiece; of the very many images I've seen of this object, yours really does get very close to the eyepiece experience of my C14, so well done with that .

[Rodd]

Just for kicks I decided to Hubble-ize the image.  This is as close to resolved throughout that I can achieve

[geoflewis]

2 minutes ago, Rodd said:

This is as close to resolved throughout that I can achieve

Whatever Hubble-ize means, the result is suberb 🙂

[Rodd]

17 minutes ago, geoflewis said:

Whatever Hubble-ize means, the result is suberb 🙂

Thanks Geof.  If you look at Hubble images of globular clusters, you'll note there is space between all stars even in the core.  All glare is gone and the star field is fully resolved.  I find it amazing.  I don't much care for Hubble stars, but its resolving power is incredible.  For this image, I tried to accomplish the impossible--to emulate Hubble (Hubble-ize).

[geoflewis]

18 minutes ago, Rodd said:

If you look at Hubble images of globular clusters, you'll note there is space between all stars even in the core.  All glare is gone and the star field is fully resolved.

18 minutes ago, Rodd said:

For this image, I tried to accomplish the impossible--to emulate Hubble

Well you got pretty darned close IMHO

 

[Rodd]

6 minutes ago, geoflewis said:

Well you got pretty darned close IMHO

Much appreciated.  I suppose it is a game of horse shoes, in a sense.  perfection is not attainable.  Only ever shorting increments of "closer"