A 45 panel 100hr plus Samyang135ASi2600 mosaic of the Bottom of Cepheus and Most of Cygnus including Elephants Trunk, Flying Bat, Lion, Seahorse, ClamShell, North America, Pelican, Propellor, Butterfly, Crescent, Tulip and Veil.

[Laurin Dave]

23 hours ago, geeklee said:

What a project and final result Dave - fantastic work, no wonder you're pleased with the final outcome.

This is a perfect size to enjoy the wider views and still showing detail viewing at full resolution.

I've seen you mention previously how much you value this tool.  Does APP do anything equivalent with it's LNC or is that an unrelated process?  I primarily work within APP pre-processing but will think about opening up the documentation for NSG and investigating further.

Lee

I'm no expert on this but I believe the objectiive of LNC is fundamentally the same as NSG....  I've no idea how LNC works but I do know that NSG uses stellar photometry to normalise the images and the author (who is a member of my astro club) has convinced me that it is more accurate and my quite extensive testing against PI's LN and APPP LNC has convinced me that it gives better  results for individual frames.  However when it comes to Mosaics, I have found that APPs registration and integration using LNC and multi band blending is far superioir to anything I can get Pixinsight to do..   Note that my input for mosaics are cropped, flattend and BlurExterminated stacks..  

Dave

[Laurin Dave]

4 hours ago, Ouroboros said:

I have to say I prefer the starless version. I wonder if there’s a compromise ie starless plus’s brightest stars only. 

Maybe..  something like this perhaps  ..  

 

[geeklee]

1 hour ago, Laurin Dave said:

has convinced me that it is more accurate and my quite extensive testing against PI's LN and APPP LNC has convinced me that it gives better  results for individual frames. 

Thanks for the additional info Dave.

1 hour ago, Laurin Dave said:

Note that my input for mosaics are cropped, flattend and BlurExterminated stacks..  

Agreed on the MBB in APP - I've found it excellent for mosaics and echo your workflow above.  I've found greater success bringing each master panel into PI and running those processes above before building the mosaic in APP afterwards.

[Ouroboros]

4 hours ago, Laurin Dave said:

Maybe..  something like this perhaps  ..  

 

Kinda works I think. 

[ollypenrice]

21 hours ago, Elp said:

AI nonsense can go in the bin.

 

For you, perhaps, but for me it can perform small cosmetic repairs which remove distractions. We can agree to disagree.

9 hours ago, Laurin Dave said:

  its where I've badly blended the IC1396 plusHa image in ...   

Dave

Rodd on here asked PI if they had any plans to facilitate the blending of high res data into widefield and was told that this was painting and no part of PI. I think this is, if nothing else, a simple minded response. I think it is a deeply satisfying form of photographic craftsmanship, one I have enjoyed developing over more than ten years and one which I think gives exciting results.

My latest approach is to blend the high res (or NB) data after star removal and using Photoshop's blend mode lighten. I find that this offers an easy way to avoid joints being visible. I make sure that all stars in the image come from the same source - the widefield.

Olly

[Stuart1971]

23 hours ago, Elp said:

AI nonsense can go in the bin.

Personally I find defects in images perfectly fine, it shows it was conducted by a person, from Earth.

The image is outstanding as it is.

Spoken by a true non PixInsight user who can’t use some of the  superb AI tools 😂😂

[Stuart1971]

On 19/08/2024 at 15:12, Laurin Dave said:

As the title suggests this is a rather large mosaic which started in the fall of 2022 as a relatively small four panel one around NGC7000 and Sadr but which just grew and grew through 2023 and 2024 to date.    As presented it is composed of 45 panels of between 2 and 4 hours each taken with a Samyang135ASI2600mc (initially at f2.8 but this year after some tweaking at f2).  Around Sadr,  the Flying Bat and Elephants Trunk Ha from the same setup but with a duo band LExt filter has been blended in to enhance the structure, also within the Flying bat Oiii has been added to show the Squid and the Cocoon has been enhanced with some Hi Res data from my trip to Les Granges 5 years ago @ollypenrice. Apart from that it is purely colour.   Proccessing was to say the least a bit complicated with much trial and error.  My workflow evolved to this which seems to me to have been succesful in acheiving a reasonably seamless mosaic:   a) calibrate, cosmetic correct and register in Pixinsight,   b) run the subs through Normalise Scale Gradient,  by normalising every sub to "the best one" this helps keep the gradient in the final stack as simple as possible and easier to remove - very useful in my Bortle 4/5,  c) Stack the normailsed subs in Pixinsight,  d) crop the edges of the Stack,  e) Run Blur Exterminator to correct the corner stars, f) Remove the stars using Star Exterminator, g) remove the gradient from the Starless Image, h) Using Pixelmath subtract the gradient produced in Step "g)"  from the non Star Removed  Stack , h) Convert ak the Stacks to TIf or FITS format, i) Load the Stacks in to Astro Pixel Processor, j) Set Registration paramaters to Mosaic, Integration paramaters to MultiBandBlendingand Local Normalisation Correction, k) Press go and come back 36hours later!

The resulting image popped out and was 25000x25000!  Cropped to suit then initial stretching in Pixinsight using GHS and HT, Stars removed using StarExterminator, further stretching in Pi, save background and Stars as TIFS, move to Photoshop for colour enhancement a minor amount of work on the seams and re-screening of the Stars back in produced the image below (presented at 50pc 10000 x5000).   

This area of space contains 173 named stars, 2 Messier objects M29 and M39, 69 NGC/IC, 136 Planetary nebulae, 270 LBN, 331 LDN, 14VdB, 32 Sharpless and 61 Barnard Objects!  (According to Pixinsight)

Thanks for looking

Dave

and annotated at 25%

 

Incredible image and huge patience….well done indeed…👏🏻👏🏻

[tomato]

I’m probably in the minority but I’m not  keen on the brighter stars only image, for me it just emphasises that most of the stars are missing. I’m not a fan of starless images either, but I prefer that to the brighter stars only version.

Dave, I recall how a few years back you worked wonders on my 12 panel CCD mosaic of M31 with the PI Photometric Mosaic script, after I was struggling with APP’s mosaic tools, it’s fascinating how image processing techniques continue to evolve.

With regard to the AI tools, I understand why some imagers have their reservations, but I’m OK with ‘em, a lot of my images would be in the bin without RC’s contributions.

 

[Elp]

3 minutes ago, Stuart1971 said:

Spoken by a true non PixInsight user who can’t use some of the  superb AI tools 😂😂

There's the phrase that comes to mind "a bad workman blames his tools". Siril has most of the necessary things built in and is free, together with good PS skills (or general art skills which are transferrable) PI isn't essential, a lot of people do without (what did we all do before it?).

I see the benefit of RC tools etc, it can also speed up workflow, I just don't think they're needed to create an artistic, subjective piece of work which is usually not to everyone's tastes. If we were making critical scientific measurements, that's a different matter.

In this case, it's an artistic piece of work cataloguing a large portion of sky, and it succeeds immensely, more so because of the time and effort that's been invested into it, a feat of which is independent of any software utilised.

[WolfieGlos]

6 minutes ago, tomato said:

I’m probably in the minority but I’m not  keen on the brighter stars only image, for me it just emphasises that most of the stars are missing. I’m not a fan of starless images either, but I prefer that to the brighter stars only version.

I'm there in the minority too, I like to see the stars in an image and not so much removed. Maybe a tad reduced from the original in the OP, but not by much at all, and not enough for me to make comment previously  

I like to see starless images when I'm processing and to see what's actually there; for me the challenge is then to show them both in a single image. I find a run of BlurXT on default settings is perfect for what I want.

[Stuart1971]

19 minutes ago, Elp said:

There's the phrase that comes to mind "a bad workman blames his tools". Siril has most of the necessary things built in and is free, together with good PS skills (or general art skills which are transferrable) PI isn't essential, a lot of people do without (what did we all do before it?).

I see the benefit of RC tools etc, it can also speed up workflow, I just don't think they're needed to create an artistic, subjective piece of work which is usually not to everyone's tastes. If we were making critical scientific measurements, that's a different matter.

In this case, it's an artistic piece of work cataloguing a large portion of sky, and it succeeds immensely, more so because of the time and effort that's been invested into it, a feat of which is independent of any software utilised.

Blah blah blah…..😂😂😂

[Laurin Dave]

8 hours ago, ollypenrice said:

For you, perhaps, but for me it can perform small cosmetic repairs which remove distractions. We can agree to disagree.

Rodd on here asked PI if they had any plans to facilitate the blending of high res data into widefield and was told that this was painting and no part of PI. I think this is, if nothing else, a simple minded response. I think it is a deeply satisfying form of photographic craftsmanship, one I have enjoyed developing over more than ten years and one which I think gives exciting results.

My latest approach is to blend the high res (or NB) data after star removal and using Photoshop's blend mode lighten. I find that this offers an easy way to avoid joints being visible. I make sure that all stars in the image come from the same source - the widefield.

Olly

Hi Olly,   I generally do much the same for blending, in this case I had already fully processed certain areas, in particular this one where I had gone much deeper and added a lot of extra higher res data,  but I blended it in with the star layer in place so didnt notice the join (lasso and feather method)

Dave

 

 

[Laurin Dave]

For those of you wondering what the image would look like if it was processed witgh the stars in ... here it is ...   BXT was run only on the individual panels and "Correct Only" ...  the Mosaic was then stretched with a combination of Global Hyperbolic Stretch (to my mind a neat PI way of doing Curves) Histogram Transformation and Masked Stretch...  Nothing else..   Amongst other things I think this shows just how good APP is at Mosaics..   I think for me the value of Star Exterminator is more that it allows you to sort out and enhance the background rather than de-emphasise the stars (masked Stretch has always been able to do that)....After all without the Stars there really would be nothing to see... and we wouldn't be here anyway!

Dave

Edited August 22 by Laurin Dave

[Laurin Dave]

8 hours ago, tomato said:

Dave, I recall how a few years back you worked wonders on my 12 panel CCD mosaic of M31 with the PI Photometric Mosaic script, after I was struggling with APP’s mosaic tools, it’s fascinating how image processing techniques continue to evolve.

Inded Steve I remember that, it was when I was doing testing for the author of the NSG and PMM scripts (John Murphy) ...   He is currently in the process of creating a stand alone pre-processing front end incorporating NSG.  Quite an epic undertaking!

Dave 

[scotty1]

Not sure how many of these you've completed, but it could make a nice picture book some time. 

🤩

[Knight of Clear Skies]

Well, that's a bit special! Sterling effort.

[Daniel-K]

😍 HOLY MOLY