Why are my stars so big? NGC891

[Chris49]

 

This is NGC891 taken as 19 subs of 90sec on 3rd Dec. using EdgeHD8 at 2032mm attached through the #93644 T adaptor to an SX694C Pro OSC, then stacked in DSS.

 

Why are the stars so big? SG Pro gives the HFR values between 3.5 and 6 - I've tried improving focus, but can't get it better.  I've used a Bahtinov mask and I've used the autofocus routine in SG Pro, but to no avail.

Is it that I need a UV/IR filter?  If so, where in the image chain can I mount it - there don't seem to be any convenient 1.25" or 2" threads to mount it with my setup.

Guiding with PHD2 gives me about 1 arcsec or better, but I have some flexure issues.

As you will have noted, I have minimal skill at processing!

I would be very pleased for any pointers as to what is happening here - don't hold back to spare my feelings, I know there is a long way to go!

Many thanks

Chris

[Davey-T]

Stars look slightly misshapen, may be flexure as you say, also a few hot pixels, if this is straight out of DSS it's not bad, stars look a bit over exposed but lots can be done in processing to correct stuff.

Stars can be shrunk, coloured, removed and replaced after sharpening the galaxy, lots of endless fun and frustration to be had and more data always helps.

Dave 

[Star101]

Could be seeing? Or wind? Or even, dare I say it, Collimation ?

They are not too bad looking at my image taken about a year ago.

 

Maybe the stars are just that large in that area of the sky.

Dave

[Chris49]

10 minutes ago, Davey-T said:

Stars look slightly misshapen, may be flexure as you say, also a few hot pixels, if this is straight out of DSS it's not bad, stars look a bit over exposed but lots can be done in processing to correct  stuff.

Thanks Dave, I did give it a bit of curves stretching in Photoshop and I had to suppress the hot pixels by hand - no dark frames - clearly I missed a few.

Any thoughts on the need for a filter?

Chris

[Star101]

I would use a LP filter no matter what.

In PixInsight, if you have it, I would use Noise Reduction / SCNR ( Green) to reduce the green cast on the image. Not sure about PShotoshop.

Check the subs and remove any that don't have perfect round stars. Taking a closer look, I feel this image was taken on a windy night.

 

Dave

[Chris49]

6 minutes ago, Star101 said:

They are not too bad looking at my image taken about a year ago.

Thanks Dave.

Presumably that is with your C11?  I've never used a refractor, but people seem to get HFR of ~1.2 with those.  Is 3.5 to 6 all I can expect with a reflector?

[vlaiv]

You are sampling at 0.46"/px and you have guide RMS error of about 1".

These are main reasons why your stars appear large. They are not in fact large, they are "average sized" stars.  Most people will get such stars in their images. Problem in your case is that you have huge pixel scale and inappropriate guiding for such pixel scale.

While these stars look big:

Once you do proper matching of resolution to setup (guiding performance and the rest), you'll get much better looking stars

As you can see - now stars look as they should.

[Star101]

2 minutes ago, Chris49 said:

Thanks Dave.

Presumably that is with your C11?  I've never used a refractor, but people seem to get HFR of ~1.2 with those.  Is 3.5 to 6 all I can expect with a reflector?

Yes, that was with the C11. Stars do look more bloated on reflectors. Personally, I focus with the Bat mask then check the stars and manually tweak the focus to get the small background stars to come into view. Then I am happy I have good focus.

[Chris49]

3 minutes ago, Star101 said:

Check the subs and remove any that don't have perfect round stars. Taking a closer look, I feel this image was taken on a windy night.

This was 19 out of 30, so a lot were thrown away.  I haven't splashed out on Pixinsight yet, I am trying to get something worth processing first!

I have made big changes to the setup - moving the #51611 guide scope on to a top bar from the normal finderscope position.  That should help balance.  Waiting for it to stop raining here!

Thanks for your thoughts.

Chris

[vlaiv]

Just now, Star101 said:

Stars do look more bloated on reflectors.

I don't think this is necessarily true, it is down to matching pixel scale, guide performance and seeing conditions. In fact due to aperture size, I constantly find that smaller refractors have larger stars (when resolution and rest is matched).

SCTs have some issues, and I personally don't like how stars look because of that - regular SCTs have a bit of spherical aberration that depends on focus position. Because focusing is done by moving mirrors - this introduces a bit of spherical aberration because mirrors are not separated with optimal distance.

I'm not sure if Edge models suffer the same - they should be better corrected.

[Chris49]

9 minutes ago, vlaiv said:

it is down to matching pixel scale, guide performance and seeing conditions

A lot to think about here.  Clearly I will try to find a place to put the filter and concentrate on improving the guiding.  Also maybe I should include the #94242 0.7x reducer - I was trying without because it is the tiny galaxies that I'm really after.

Thank you for your advice.

Chris

[carastro]

There may also be an element of overstretching the stars in processing as you say your processing skills are minimal.  What software are you using for post-processing?

Plus as Vlaiv says you could be undersampling, I found I got larger stars for that reason once.  

Carole 

Edited December 19, 2019 by carastro

[vlaiv]

2 minutes ago, Chris49 said:

A lot to think about here.  Clearly I will try to find a place to put the filter and concentrate on improving the guiding.  Also maybe I should include the #94242 0.7x reducer - I was trying without because it is the tiny galaxies that I'm really after.

Thank you for your advice.

Chris

You have hit major obstacle (and hit it rather hard) - there is limit to how "close" in we can get. After you cross that line - you just end up with zoomed in image - but blurry and without detail. Bloated stars are obvious consequence of that.

If you are after maximum magnification and as close in as you can get for galaxies or other small object (like planetaries), you want to do 3 things.

Well - you want to do two things, and wait for third - as there is noting you can do about it, except moving to Atacama or similar.

Do whatever it takes to get your guiding sorted and that means guide RMS at 0.5" or below. Make your sampling rate somewhere between 1 and 1.2"/px and of course - wait for night of good seeing. Actually that last one can be helped - avoid anything that can mess up seeing locally (read on how to optimize planetary viewing) and work on targets when they are highest in the sky and / or best positioned.

In order to get to certain resolution with EdgeHd8" and that color camera, you need to do two things - add reducer and debayer using super pixel mode. That will put you in 1.32"/px and that is quite ok to start with for high resolution work.

[Chris49]

3 minutes ago, carastro said:

What software are you using for post-processing?

Stacked in DSS, stretched a bit in Photoshop - I may well have clipped off the brighter stars.

Chris

[carastro]

OK, in Photoshop. in levels, you have a black point, a grey point and a white point.  I never move the white point, it just causes enlarged burnt out stars.  Even the grey slider can enlarge the stars a bit, so need to use that carefully.

If you are using curves, I always anchor the white point down (on the diagonal line = click on the white stars and watch where the mini box pops up and then click on that spot) then when you stretch in Curves it will help stop the white being stretched.

HTH

Carole 

 

 

Edited December 19, 2019 by carastro

[Chris49]

5 minutes ago, vlaiv said:

Well - you want to do two things, and wait for third - as there is noting you can do about it, except moving to Atacama or similar.

Do whatever it takes to get your guiding sorted and that means guide RMS at 0.5" or below. Make your sampling rate somewhere between 1 and 1.2"/px and of course - wait for night of good seeing. Actually that last one can be helped - avoid anything that can mess up seeing locally (read on how to optimize planetary viewing) and work on targets when they are highest in the sky and / or best positioned.

In order to get to certain resolution with EdgeHd8" and that color camera, you need to do two things - add reducer and debayer using super pixel mode. That will put you in 1.32"/px and that is quite ok to start with for high resolution work.

Thanks Vlaiv, that's very helpful. I guess I have been unrealistic up to now.  I will work towards what you suggest.

Chris

[Chris49]

6 minutes ago, carastro said:

OK, in Photoshop. in levels, you have a black point, a grey point and a white point.  I never move the white point, it just causes enlarged burnt out stars.  Even the grey slider can enlarge the stars a bit, so need to use that carefully.

If you are using curves, I always anchor the white point down (on the diagonal line = click on the white stars and watch where the mini box pops up and then click on that spot) then when you stretch in Curves it will help stop the white being stretched.

Thank you Carole.  I will be more careful next time.  So far I have had very little idea what I am trying to do and even less about how to do it!  It's all part of the learning curve. 

Chris

[JamesF]

There are techniques in PS for reducing star size, too.  I'm trying to get my head around them myself at the moment (I searched for "photoshop star reduction" and started from there).  Obviously keeping the star size small to start with is better, but needs must...

It's not clear from the specs for the camera on the SX site whether the optical window rejects IR/UV or not.  If it doesn't then it may be worth looking at how you might fit an IR/UV filter.  Or, if you're working in a fairly light-polluted area perhaps it is worth looking at an LP filter as has already been suggested.

James

Edited December 19, 2019 by JamesF

[Davey-T]

You can "fix" eggy stars in P'Shop by copying the image onto itself and using blend mode darken nudge the layer with arrow keys to improve the stars.

Dave

[Chris49]

5 minutes ago, JamesF said:

It's not clear from the specs for the camera on the SX site whether the optical window rejects IR/UV or not.  If it doesn't then it may be worth looking at how you might fit an IR/UV filter.  Or, if you're working in a fairly light-polluted area perhaps it is worth looking at an LP filter as has already been suggested.

I don't think it does reject UV, hence my worry that UV might be out of focus and lead to blobby stars.  I thought that was more a problem with refractors than reflectors though, although there is the corrector plate I suppose.

I do have an LP filter to try, but although Clear Outside says I have Bortle 4, I find that hard to believe!

I will search for PS star reduction as you suggest.

Thanks

Chris

[ollypenrice]

SCTs can deliver astonishing planetary detail in fast frame imaging but are notorious for giving large stars in deep sky work. I don't know why that is, but the evidence is overwhelming. 

This target is surrounded by bright stars, which doesn't help. You will need excellent guiding, excellent focus and excellent seeing to hold them down.

And then, if you're looking at good images on the net, you'll be looking at those produced by experienced image processing experts who have a battery of techniques for holding down stars. In this image I wouldn't even bother holding most of them down, I'd just combine two different stretches, one for the galaxy and one for the stars. This is really for a different discussion but my point is that processing plays a big role in star control and there is no single best way to do it.

Olly

[carastro]

Quote

There are techniques in PS for reducing star size, too.  

I recently uploaded this video on reducing star size, it uses a technique that MartinB posted up a few years back which i have found invaluable.  If the video starts in the middle just slide it back to the beginning.  I don;t know why it does that.

Carole 

Edited December 19, 2019 by carastro

[Chris49]

6 minutes ago, Davey-T said:

You can "fix" eggy stars in P'Shop by copying the image onto itself and using blend mode darken nudge the layer with arrow keys to improve the stars.

That sounds a really clever technique.  There's a lot to this astro stuff when you get into it!

Thanks

Chris

[JamesF]

2 minutes ago, carastro said:

I recently uploaded this video on reducing star size, it uses a technique that MartinB posted up a few years back which i have found invaluable.

Carole 

Thank you for that, Carole (and MartinB :).  I shall watch it later this evening if I don't fall asleep first.

James

[Davey-T]

1 minute ago, Chris49 said:

That sounds a really clever technique.  There's a lot to this astro stuff when you get into it!

Thanks

Chris

Bit of a bodge really, bear in mind that some "experts" spend tens of hours processing their images and they've got excellent data to start with, as Olly said if you look for images of the same target on the net you may get disheartened at not being able to match them, I gave up trying years ago 

Dave