M27 Two hours in Astro Twilight

[cuivenion]

Hi, This is M27 taken over three nights during Astro Twilight, mid to late May and early June. It totals 2 hours and 20 minutes of exposure.

Equipment:

Skywatcher HEQ5, Skywatcher 200p, ZWO ASI224. Imaged at 0.77" (1000mm focal length)

Software:

Captured with Sharpcap. Stacked with DeepSkyStacker and processed with Pixinsight

1055 x 8 second unguided light frames.

No darks, flat or bias. I have taken darks but I have decided against using them as a sigma stacking process gets rid of the hot pixels, and DBE in Pixinsight along with cropping seems to have taken care of the slight amp glow.

I think I'm going to add more lights to this and reprocess and add to this thread so feel free to comment as I'm still learning Pixinsight.

 

[MikeODay]

Very nice image with great colour and detail.

I am truely surprised that you can pull so much information out of the background with just 8 sec exposures.  Very impressive!

[cuivenion]

Thanks Mike!

[Demonperformer]

Yes excellent for 8s subs. The ZWO delivers again.

[ollypenrice]

That's good going for the time of year from the north of England! Lots to like.

You do have a slight colour offset in the stars right across the image. Many look a tad blue on the left limb and a tad red on the right.  Is this OSC or mono? If it's OSC it isn't obvious how it might arise.

Olly

[cuivenion]

Thanks guys! Olly I did perform colour calibration in the processing or do you mean an error in the initial stacking? Thanks for the info.

[cuivenion]

Hi, it is an OSC camera. I'll reprocess soon and see if the early stages show the offset.

[ollypenrice]

11 minutes ago, cuivenion said:

Hi, it is an OSC camera. I'll reprocess soon and see if the early stages show the offset.

If it's OSC it can't be channel alignment so it must be of optical origin. I don't know if slight mis-collimation might introduce this or if it might come from glass in the light path (Coma corrector? Filter window? The Bayer filters themselves?) Or what about a bit of tilt in the system? I'm not at all sure.

Olly

 

[cuivenion]

Hi, there is only a optical window to protect the sensor from dust, it's AR glass. No coma corrector needed fortunately with the asi224 due to it's small FOV. I do collimate before every session with a cheshire but I couldn't swear I'm an expert, it looks OK to me.

It could be possible there's tilt, the camera is just put in like an eyepiece into the 1.25" holder and tightened with the 2 thumbscrews. I probably need to to get a T2 extension tube so I can connect by screwing directly onto the focuser.

[Sunshine]

Great image! i don't know why its not called the Fisheye nebula, sorry i mean the dead fisheye nebula, don't buy fish with eyes like that lol.

[cuivenion]

Thanks sunshine!

[cuivenion]

2 hours ago, ollypenrice said:

That's good going for the time of year from the north of England! Lots to like.

You do have a slight colour offset in the stars right across the image. Many look a tad blue on the left limb and a tad red on the right.  Is this OSC or mono? If it's OSC it isn't obvious how it might arise.

Olly

I've been searching and found this:

https://pixinsight.com/forum/index.php?topic=10513.0

It was only at 35 degrees when I started imaging. It looks like this could be a factor. 

[ollypenrice]

24 minutes ago, cuivenion said:

I've been searching and found this:

https://pixinsight.com/forum/index.php?topic=10513.0

It was only at 35 degrees when I started imaging. It looks like this could be a factor. 

That's a good discussion. While I agree with Alex W that atmospheric effects do vary with altitude, I wonder why you would get a systematic shift of red to one side and blue to the other? If you look at an early and late sub, stretching both, do you see the same thing? Does it get worse?

Mike's suggestion of aligning separated blue and red channels onto green is logical, green being mid spectrum. It presumably works for him and may work for you. However, on the rare occasions when I have channel alignment problems I find, for some reason, that it's best to make red the definitive channel. Be prepared to experiment flexibly.

Olly

[cuivenion]

The lights were all from different nights as well so that could complicate things.

[HunterHarling]

Nice image.

I see this effect on the stars when my telescope is out of collimation. Might be worth checking...

[MikeODay]

14 hours ago, ollypenrice said:

That's good going for the time of year from the north of England! Lots to like.

You do have a slight colour offset in the stars right across the image. Many look a tad blue on the left limb and a tad red on the right.  Is this OSC or mono? If it's OSC it isn't obvious how it might arise.

Olly

My images suffer a little from the same or similar issue.  I had put it down to atmospheric dispersion.  Is that a possibility?

Typically this is what I see in my images ( if it is not atmospheric dispersion I'd love to know what it is in case I can do anything about it  )

 

12 hours ago, cuivenion said:

I've been searching and found this:

https://pixinsight.com/forum/index.php?topic=10513.0

It was only at 35 degrees when I started imaging. It looks like this could be a factor. 

The link did not work for me.

  

11 hours ago, ollypenrice said:

That's a good discussion. While I agree with Alex W that atmospheric effects do vary with altitude, I wonder why you would get a systematic shift of red to one side and blue to the other? If you look at an early and late sub, stretching both, do you see the same thing? Does it get worse?

Mike's suggestion of aligning separated blue and red channels onto green is logical, green being mid spectrum. It presumably works for him and may work for you. However, on the rare occasions when I have channel alignment problems I find, for some reason, that it's best to make red the definitive channel. Be prepared to experiment flexibly.

Olly

 

In my case, splitting the OSC into channels and aligning them on one colour did not reduce the issue for me -  it only shifted the channels by a small fraction of a pixel, whereas the colour shift in the stars is maybe 3 to 5 pixels on very bright stars.

[Demonperformer]

To my uneducated eye, that looks like the result of passing through an uncorrected frac. Could it be a problem with the coma corrector?

[MikeODay]

43 minutes ago, Demonperformer said:

To my uneducated eye, that looks like the result of passing through an uncorrected frac. Could it be a problem with the coma corrector?

It could be.  But I guess I would have expected more distortion of the star shape ( at least the examples on the internet of chromatic aberration seem to show elongated stars - or maybe they are just extreme examples ).

I suspect my setup suffers from a number of small problems so it is hard to isolate just one main cause - I know it suffers from minor/moderate astigmatism.

 

16 hours ago, ollypenrice said:

That's a good discussion. While I agree with Alex W that atmospheric effects do vary with altitude, I wonder why you would get a systematic shift of red to one side and blue to the other? If you look at an early and late sub, stretching both, do you see the same thing? Does it get worse?

Mike's suggestion of aligning separated blue and red channels onto green is logical, green being mid spectrum. It presumably works for him and may work for you. However, on the rare occasions when I have channel alignment problems I find, for some reason, that it's best to make red the definitive channel. Be prepared to experiment flexibly.

Olly

5 hours ago, MikeODay said:

In my case, splitting the OSC into channels and aligning them on one colour did not reduce the issue for me -  it only shifted the channels by a small fraction of a pixel, whereas the colour shift in the stars is maybe 3 to 5 pixels on very bright stars.

I take that back

A new test has revealed that splitting the channels just after calibration, aligning all the channels to one channel ( one of the green in my case ), integrating the R, G and B sets separately and then recombining them into an RGB image has ( at least in my case ) significantly reduced the colour aberration ( regardless of its cause ).

old:  

New:

 

These are 100% crops from the corner of an image, enlarged 3x to make them easier to see and saturation boosted way too high to more clearly show the colour variation within the stars.

The second ( processed as individual channels ) is much improved.   ( Still some issues - the main one being red splotches at the edge of the stars at about 100 and 280 degrees and smaller blue splotches at about 10 and 190 degrees ( these angles co-respond with the angle of the astigmatism in my system )).

The upshot of all that is that it may well be worth trying this approach to see if it can reduce the spread of colour across stars.

[ollypenrice]

Gosh yes, that's a vast improvement. It should make the image look significantly crisper. I do have colour aberrations sometimes, very like these, when I over-do Noel's Action for increasing star colour, so I guess it can be latent in many images.

The other thing that comes to mind is the anti-aliasing filter used in OSC stacking algorithms. In AstroArt its parameters are adjustable. I don't know about PI.

Olly

[cuivenion]

Thanks for the great input guys. Mike I'll be trying that today that's a great improvement.

Demon there was no CC used although it's possible the AR glass in front of the sensor has a problem.

Hunter I have noticed that my 200p can slightly lose collimation between sessions whereas my 130pds tends to stay rock solid. It's the primary mirror that moves. Maybe the mirror is shifting during the session.

Olly I'll look into that, although after moving to PI for image integration I've started using DSS again as I found after spending hours manually weighting, aligning, debayering etc, I was getting the same results if I was just running it through DSS which does it automatically and quicker.

[cuivenion]

On 06/06/2018 at 05:28, MikeODay said:

It could be.  But I guess I would have expected more distortion of the star shape ( at least the examples on the internet of chromatic aberration seem to show elongated stars - or maybe they are just extreme examples ).

I suspect my setup suffers from a number of small problems so it is hard to isolate just one main cause - I know it suffers from minor/moderate astigmatism.

 

I take that back

A new test has revealed that splitting the channels just after calibration, aligning all the channels to one channel ( one of the green in my case ), integrating the R, G and B sets separately and then recombining them into an RGB image has ( at least in my case ) significantly reduced the colour aberration ( regardless of its cause ).

These are 100% crops from the corner of an image, enlarged 3x to make them easier to see and saturation boosted way too high to more clearly show the colour variation within the stars.

The second ( processed as individual channels ) is much improved.   ( Still some issues - the main one being red splotches at the edge of the stars at about 100 and 280 degrees and smaller blue splotches at about 10 and 190 degrees ( these angles co-respond with the angle of the astigmatism in my system )).

The upshot of all that is that it may well be worth trying this approach to see if it can reduce the spread of colour across stars.

I have gained a liitle improvement. I think it may be a case of finding the right file to align with. The effect is still there unfortunately. I wasn't using any calibration files. I split all the channels and then aligned and stacked seperately used a green file as a reference for all the red, green and blue files. I then combined the red, blue and green integrated images, is that correct? One thing I've noticed is its only the bright whiter stars suffering from this effect. The dimmer red stars are fine.

The brighter stars also have their shape effected by the focuser cutting into the light path, and thats where the blue effect occurs. The focuser is undarkened. Could this be the culprit?

 

[Xiga]

Just seeing this discussion now. As someone who does all his imaging through a basic SW 80ED, I regularly have stars that are coloured differently on both sides. But just recently, I came across a setting in AstroPixelProcessor which claims to significantly reduce chromatic aberration so I thought I'd give it a go. It's basically nothing more than a tickbox so takes zero effort (always a good thing!). I can't remember exactly where it is, I'm at work so can't check, but I think it's somewhere near the bottom of the Calibration tab. I only tested it on one image, but I have to say I thought it really did make a noticeable difference, so I intend to use it now on all of my OSC images.

There's a free trial of APP available so no reason not to download it and give it a go. It certainly sounds a lot easier than the workflow above.

[cuivenion]

Thanks, I'll give it a look.

[MikeODay]

12 hours ago, cuivenion said:

I have gained a liitle improvement. I think it may be a case of finding the right file to align with. The effect is still there unfortunately. I wasn't using any calibration files. I split all the channels and then aligned and stacked seperately used a green file as a reference for all the red, green and blue files. I then combined the red, blue and green integrated images, is that correct? One thing I've noticed is its only the bright whiter stars suffering from this effect. The dimmer red stars are fine.

The brighter stars also have their shape effected by the focuser cutting into the light path, and thats where the blue effect occurs. The focuser is undarkened. Could this be the culprit?

 

The process you describe sounds right.

By the way, In the past when I tried stacking the different channels it did not make much a difference.  In the recent experiment I turned on the option of distortion correction.  I suspect that this may have helped.

[cuivenion]

No probs. I was thinking of trying that as I've seen it mentioned elsewhere. Hopefully it'll make a difference. Thanks for the help.