NGC 6946 (Fireworks Galaxy)

[NigeB]

Hello All

I managed to get what I thought were some reasonably good subs of the Fireworks Galaxy last week and have been trying to get a good result in Pixinsight. Results shown below. I used the processing workflow described by @MrsGnomus in an earlier thread, with some additional adjustments. I've added some H-alpha into the red channel. 

Exposure details: 15 x 10 min L, 10 x 10 min R, 11 x 10 min G, 10 x 10 min B, 8 x 10 min H-alpha added to R channel (0.9 R + 0.4 Ha). Taken with the TOA-150 and 1.5 extender (so running at f/11). 

On closer inspection the exposures aren't as good as I'd hoped - the star shapes are consistently elongated. Another set of images I took of M57 the week beforehand, have nice round stars. No change in configuration between the two targets, so I'm wondering if the higher declination of NGC 6946 has something to do with it - maybe my PHD2 calibration wasn't good.

Comments & criticism welcome - as always, I'm happy to provide raw data or stacked subs if anyone else feels like having a go!

Regards

Nigel

 

 

 

 

[Spaceade]

Great image. Didn't spot the slightly-elongated stars until after i read the text. It's fine

[NigeB]

2 hours ago, Spaceade said:

Great image. Didn't spot the slightly-elongated stars until after i read the text. It's fine 

Thanks Spaceade!

[Petergoodhew]

It's a lovely image.  I'm struggling to see the star elongation too. One possible explanation for it could be wind conditions, given the long OTA of a 150 refractor, increased exposure from the higher elevation, and high sensitivity resulting from being F11.

Personally I would like to see a little more red, so would add the Ha to 1.0 R rather than 0.9 R.

[NigeB]

Hi Peter,

Many thanks for your comments; I'm pleased the elongation doesn't look too obvious. Once I saw it, I couldn't stop seeing it... 

I did wonder about wind, but it was a pretty calm sequence of nights. Good point about the extra sensitivity at F/11. I've gone back and looked at the subs again; the extent and direction of trailing seems consistent in all of them. I've checked the orientation of the long axis of the stars, and it seems to be very well aligned with the RA axis, so I'm still thinking that a poor PHD2 calibration before moving to this high Dec could be to blame. I might have another go with a forced re-calibration on something near the celestial equator and try a couple of subs on the same target to see if it behaves any differently.

Thanks for the suggestion about changing the Ha balance; that's very useful feedback. It's the first time I've tried adding Ha to an LRGB image and I went with a very light touch. I'll revisit at the weekend and post an update.

Best Regards

Nigel

 

[wornish]

Really nice image - well done.

I honestly don't understand this fixation about star elongation I can't see any in your image, which particular star is bothering people?  and more importantly why?

I would really like to have a go at processing the data if you are happy to share.

[Stub Mandrel]

Well done Nige, much better than my feeble efforts on this target for sure!

[ollypenrice]

It's a good result. Greens look a bit high to me. Dead easy fix with SCNR green though.

Olly

[NigeB]

Hi Dave, Neil, Olly,

Many thanks for your positive comments - it's good to know that this is heading in the right direction and people don't find the issues too distracting. I'm going to re-process with more H-alpha and a green correction in SCNR and see how that goes - I appreciate your suggestions.

Dave, I agree the elongation isn't immediately obvious. But I knew there was something amiss when I saw the subs next morning, even before I looked closely at the individual stars. The images just didn't look quite right. I get grumpy over star shapes because I know that if all is well, then in the well corrected central portion of the FOV the stars should be perfectly circular. They're a very sensitive indicator of some types of problem, like guiding, so if the stars are slightly out of circular, then that indicates a problem which is affecting the whole image, including diffuse features.

Attached is a screenshot of a small portion of the stacked L frame not far from the field centre (all the stacked frames and individual subs show the same). I can see the issue in all of the stars here - the diameter in the horizontal axis is slightly, but noticeably, larger than in the vertical. To my eyes the zoomed view has a sense of minor but noticeable "motion blur" as if things were moving from left to right. So features like the dark lanes aren't as clear as they could be from this setup because I'm basically smearing out detail.

Anyway: yes, please do have a go at processing this - post the results (and preferably summarise your steps - it's a great way of learning technique!) Last time I did this people told me they only used the stacked and aligned frames, not the raw subs. So I've put stacked and aligned L, R, G, B and H-alpha frames (after bias, dark and flat field correction) on my Dropbox area here:

https://www.dropbox.com/sh/1kmko02tkbxqxhx/AACjT-ptgRiiVyxz7yX6nPTGa?dl=0

Let me know if you have problems accessing the files - and thanks again all.

Best Regards

Nigel

 

 

[ollypenrice]

You're right that there is a slight R/L elongation but it really is very slight. I think it's fine to try to get to the bottom of it but it would be a shame to let it spoil the image for you.

Olly

Edit: While we're being pernickety, I'd be very careful where you place your background markers in DBE because I suspect they might have been a tad too close to the galaxy. There's the tiniest hint of a dark ring around it, maybe? Or am I inventing it? ABE tends to do this on galaxy images, I've noticed.

Edited September 1, 2018 by ollypenrice

[wornish]

Thanks for the explanation of why star roundness is important.  I never thought of it that way.  Under normal viewing distances its really not noticeable in your data but the goal is a perfect image.

 Will have a go at your data when I get a chance. It has downloaded OK.

[Stub Mandrel]

You might want to try motion blur correction in Astra image.

Works well for images with low noise and small, consistent blur. I suspect it would be best to correct the subs before stacking

[ollypenrice]

There's a Photoshop fix as well. Copy layer, set top layer to blend mode darken, go to Filter-Other-Offset and experiment with a horizontal pixel offset. I went for 4 pixels and then used Edit-Fade to apply only 90% of that. It's thuggery, of course!

Olly

[Laurin Dave]

Hi Nigel,

Nice image, I had a look at your data last night whilst waiting patiently for my power supply to fail as I imaged the Fireworks galaxy in Ha. To me Dynamic psf in Pixinsight seems to show that the minor elongation gets worse on the Red from top left to bottom right and on the Green and Blue in the opposite direction, also the Blue and Green are generally more elongated. Also when combining RGB and after background neutralisation there is a pronounced  red to green gradient left to right...  So these make me wonder whether you did a meridian flip between red and green/blue and whether you have balance and polar alignment/field rotation issues, I don't know whether you are using an OAG or guidescope or are on a on a permanent pier. 

Hope this helps

Dave

[NigeB]

Wow... Some great advice and suggestions here, thank you everyone.

Olly, thanks for your further input. When I processed this first time around I did see some really strong dark rings around the galaxy, so I went back to the start to try and reduce that, but obviously not enough. I'll keep a closer eye on that. That's an interesting "thuggery" approach - and it looks pretty effective. I'll look at that though I don't have PS; I weighed up the software options and went for PI as my first processing splurge - maybe PS is in my future. In the mean time I'll look at GIMP to see if there's a similar approach.

Neil, thanks for the Astra image pointer - I was not aware of this piece of software - looks very interesting.

Dave - very useful comments; I did see the colour gradient in the stretched background though I didn't think much of it at the time since it seemed to drop below visibility after processing, but indeed, there's probably some useful information there to help track down the cause, and I missed that difference in the elongation across different colour frames. I'll need to look back at the logs and correlate the image times to see if there was a flip at that time - there certainly were flips over the nights I took the data. And it's a while since I checked alignment - it's a permanent pier mounted setup, and I use an OAG, but I've been messing about, experimenting with dual mounting telescopes and swapping them around (covered in another thread) so I'll need to give that a check. Typical - it's cloudy tonight... 

I'll re-post once I've done some more digging. 

 

Thanks once again.

 

Nigel

[wornish]

Thanks Nigel for sharing your data. Finally had a go but I think your original is better than my effort.

I didn't use the Ha data this is just LRGB processed in PI and then tweaked in PS.

 

 

Edited September 4, 2018 by wornish
further processed

[EdJP]

Hi Nigel,

I'd like to have a go at this. Can you please share your files for the Fireworks Galaxy?

Thanks,

Ed

[NigeB]

6 hours ago, EdJP said:

Hi Nigel,

I'd like to have a go at this. Can you please share your files for the Fireworks Galaxy?

Thanks,

Ed

Hi Ed,

Sure... Try this DropBox link:

https://www.dropbox.com/sh/aaioncmahrge7fi/AACgmPbGzM8Csqus5qUmUP5Oa?dl=0

In there you should find the calibrated, corrected and co-added L,R,G,B and Ha files in FITS and PixInsight .XISF format.

Best Regards

Nigel