And another M31

[tomato]

Here is M31, a 6 panel mosaic captured in a single night, 3^rd  - 4^th October.

Each panel is 40 mins Lum with the Esprit 150/0.77FR/G28300 CCD, and 40 mins RGB with the Esprit 150/QHY268c

These were combined in Pixinsight using the Photometric mosaic script, then processed in Affinty photo.

Its not terribly deep but the panels went  together OK, better than my previous M31 mosaic efforts. The gradients were easier to manage but it was a super transparent night, for a change.

[Paul M]

Wow! That's bright. I had to turn down the volume!

Then I thought about your scope. I'm not familiar with those lensy type scopes so googled it at FLO. Errr, it cost more than all my astronomical purchases over 45 years combined!!  

[tomato]

Esprits have gone up a quite bit recently, but still cheaper than a well lauded Japanese brand. One is on loan from @Tomatobro, and the other I purchased second hand.

Isn’t this why the government gave me access to my pension pot?

[Paul M]

3 minutes ago, tomato said:

Isn’t this why the government gave me access to my pension pot?

Say no more... 😎

I'm 10 years off taking my lump sum (aiming for 67  ) and by then it won't buy a decent car - if private ownership is still a thing - so maybe a new scope and mount would be on the cards...

[tomato]

Indeed, when you retire you need a hobby, swinging golf clubs never appealed.

[vlaiv]

59 minutes ago, tomato said:

Here is M31, a 6 panel mosaic captured in a single night, 3^rd  - 4^th October.

I'm struggling to understand how this image came to be? Why is it 14224 x 11407 px and how did you combine 6 panels to get that?

Why is it so terribly over sampled? Quick measurement gives about 0.7"/px, but you say you took this image with Esprit 150 with 0.77 FR - which would make it about 800mm of FL?

Pixel size needed for 0.7"/px at 800 FL is 2.7µm. Did you drizzle x2 base resolution from 8300 for some inexplicable reason?

If you look at the image in 100% - stars are huge:

This looks over sampled by factor of x6!

[tomato]

Hi Vlaiv,  I’m not at my processing PC so I can’t give you much in the way of numbers, but if I elaborate on my workflow this might help explain your observations.

Lum data was collected with the 150/FR/G2-8300 @ 1.33”/pixel, RGB data (with the OSC  camera @ 0.7”/pixel.

A single night capture necessitated a 6 panel mosaic which resulted in a “letterbox” FOV of the galaxy.

The Photometric Mosaic script requires a widefield image of the target as a reference, this is used to ensure smooth joins and minimal gradients as the panels are added. Inevitably, there were sections of background either side of the galaxy which were not covered by my mosaic panels, quite how the script deals with this I don’t know, but it returned no errors and the final image had a smooth, seamless background. 
 

The image size of the Lum mosaic was different to the RGB mosaic, these were registered in APP, using the dynamic distortion algorithm, normalised, then LRGB combined in APP. To give a more pleasing image I cropped it to the proportions of the widefield image (art intervening I’m afraid), then processed further in APP and AP.

[tomato]

I might add that I intended to software bin the image to 35% in StarTools, but alas I suspect the file size was too big as ST repeatedly crashed when trying to execute the operation.

[Laurin Dave]

That’s worked rather well Steve. Next step is to combine your other data  with it 🙂.  
 

Dave

[Ouroboros]

It is a very pleasing and bright image. I’m struggling to get my head around the processing and combining of all those files. I have enough with processing one OSC image.  

[tomato]

1 hour ago, Laurin Dave said:

That’s worked rather well Steve. Next step is to combine your other data  with it 🙂.  
 

Dave

Thanks Dave, I have been considering this, one for the next run of cloudy nights. I’m looking at @vlaiv’s comments, I’ve managed to load a LRGB file into Startools so will look at producing a version with a more sensible imaging scale.

[tomato]

I succeeded in loading separate LRGB channels into Startools, my preferred processing package, so here is a version from there, software binned 35% to hopefully address the issues noted by Vlaiv.

I note there is a theme developing with my image posts. Best to ignore the first couple or three versions I put up,  the better ones tend come along after heeding all the good advice.🙂

[Paul M]

That "looks" better to me. But vlaiv will be along soon to reburst your bubble  🤣

On a serious note, I've learned much about my own efforts from vlaiv's analyses, and I do see it as analysis rather than critique!

[wimvb]

15 minutes ago, tomato said:

the better ones tend come along after heeding all the good advice.🙂

Actually, I like the fist image better. Sure it's bright, but has more colour variation outside the bright core, showing brown colour in the dust lanes, and clear blue star clusters. Nr 2 is too "flat" for my taste.

Edited October 8, 2021 by wimvb

[tomato]

22 minutes ago, wimvb said:

Actually, I like the fist image better. Sure it's bright, but has more colour variation outside the bright core, showing brown colour in the dust lanes, and clear blue star clusters. Nr 2 is too "flat" for my taste.

I'm OK with both of 'em but  I look forward to the quantitative analysis...

[tomato]

And here is another one entirely processed in APP, no binning so fat stars at 100% scale but maybe turned the brightness down?

I can do this all night...

 

 

[vlaiv]

20 minutes ago, tomato said:

And here is another one entirely processed in APP, no binning so fat stars at 100% scale but maybe turned the brightness down?

I like this version the best - as far as processing goes.

[wimvb]

1 hour ago, tomato said:

And here is another one entirely processed in APP, no binning so fat stars at 100% scale but maybe turned the brightness down?

I can do this all night...

 

 

A brighter background? Perhaps a little less contrast in the galaxy. It makes the image softer. 👍

[Laurin Dave]

Hi Steve.. I think  your last one’s the best, background maybe looks a bit green on my screen, but very good, well done

 

Dave

[gorann]

A very nice general impression, and the last one for me too Steve. I was qurious about how well the mosaic-process worked so I could not stop myself from a bit of pixel peeping. Sorry! There are some rather odd artefacts here and there, including colour falling outside the stars. I have hard to believe that it is the fault of the Esprits and think it may have something to do with the stacking or mosaic procedure. I do all of that in PI and have not seen such artefacts there. However, doing mosaics in PI is a bit more manual (at least for me that do not know all the PI tricks).

I hope you do not mind me dissecting your image - but since also have dual rigs I am on the look out for good ways of doing mosaics.

Cheers, Göran

[wimvb]

It’s easy to clean up those artefacts with a tight star mask. Then mlt or mmt with 2 layers disabled (double clicked). If applied to chrominance, you won’t lose detail.

Edited October 9, 2021 by wimvb

[vlaiv]

Pixel peeping spreads like wildfire

 

[tomato]

I suspect the background artefacts which have distinct transition lines are due to the fact my mosaic panels did not cover the whole of the reference image, but I could be wrong. I will investigate the colour artefacts around the stars, but I have since noted that the star shapes in the corners of the Lum subs taken with the F/R are not as good as the OSC data...

As for pixel peeping, I don't mind at all, the exercise can reveal shortcomings in the processing that can be addressed. And with 145 million of them, there is no shortage of opportunities.

[tomato]

I have checked an RGB channel combination image, there are no significant colour artefacts around the stars, so it may be due to the distortion manipulation required to combine the lum channel which has a different imaging scale.

Wim, I confess I'm not familiar with your star mask fix, but there are other tools I can use.

[Rodd]

On 08/10/2021 at 14:30, Paul M said:

critique!

Yes turn the brightness way down