M31 HaLRGB-Reprocessed

[Rodd]

I have been processing this data for over a year and I finally tried something that I had previously not tried--masked stretch in PI.  While masked stretch typically results in a pretty ghastly looking image at first, I noticed that I typically overstretched my data.  Masked stretch really does a good job with the core of m31--but the rest of the galaxy is a bit flat--So  I created 2 images , used masked stretch for one and a light manual stretch for the other--I then combined the images to a much better result than I was getting with manual alone.  Of course, I did the same thing with the master lum before inserting it into the RGB image.  Here is the result.  I think it far superior than my older attempts--so i wont post the previous images.  The question I have is one of palette.  I struggle with this data to bring out the blue regions--probably need more data.  So I resorted to a palette tweak using Wndows--just a slight tweak to impart a little more blue.  But did it also impart blue to regions where blue should not be?  Please let me know which image is preferred--the first one is what I am currently calling my final image.  The second version is the one with a slight palette tweak.  I am a bit torn.  i want to say #2 because their is more blue, but my gut tells me that it isn't right.

Which is preferred? (They are the same image except for 1 palette tweak)

Final

 

Final with increased blue

 

[Kinch]

Of the two.....I prefer the top image. 👍

[Rodd]

1 minute ago, Kinch said:

Of the two.....I prefer the top image. 👍

Thanks Kinch.  I agree with you at least 50% of the time!  I waffle.  But I do expect the top one to be the more "correct" image. 

Rodd

[Kinch]

1 minute ago, Rodd said:

But I do expect the top one to be the more "correct" image. 

To be honest....spending most of my time working with NB data - I could not say which is more "correct"...... but on my monitor, the top one is more natural....if I can put it that way.

NB.....working on images for a year or more can be dangerous for your mental health!   🤔

[Rodd]

3 minutes ago, Kinch said:

working on images for a year or more can be dangerous for your mental health! 

Only with respect to broadband!  I love narrow band too.

Rodd

[Marvin Jenkins]

Rodd I love both of them. This is especially close to my heart as last night I took my first images of M31. If I can get close to your image one day I would be overjoyed.

Marvin

[Rodd]

1 minute ago, Marvin Jenkins said:

Rodd I love both of them. This is especially close to my heart as last night I took my first images of M31. If I can get close to your image one day I would be overjoyed.

Marvin

Thanks Marvin.  I remember my first successful sub....it was M31 too.  I could not figure out how to do anything (its not like there is a manual ya know?) and when I finally got guiding to work and was able to take a 5 min sub of M31, as soon as it popped up on the screen I whooped and hollered and ran inside and was overjoyed.  Those were the days!

Rodd

[geordie85]

I do prefer the 1st image. That's not to say the 2nd isn't beautiful though, it really is an outstanding image 

[Marvin Jenkins]

12 minutes ago, Rodd said:

Thanks Marvin.  I remember my first successful sub....it was M31 too.  I could not figure out how to do anything (its not like there is a manual ya know?) and when I finally got guiding to work and was able to take a 5 min sub of M31, as soon as it popped up on the screen I whooped and hollered and ran inside and was overjoyed.  Those were the days!

Rodd

I know just what you mean. Mine was three minutes which is pushing it on an NEQ5 and if you don't mind here is the end result. What do you think? Nothing done just a raw image.

[Rodd]

11 minutes ago, geordie85 said:

I do prefer the 1st image. That's not to say the 2nd isn't beautiful though, it really is an outstanding image 

Thanks Geordie!

Rodd

[Rodd]

4 minutes ago, Marvin Jenkins said:

I know just what you mean. Mine was three minutes which is pushing it on an NEQ5 and if you don't mind here is the end result. What do you think? Nothing done just a raw image.

That looks pretty good.  Hard to compare with my first as I was using a mono camera and I think it was a lum sub--could be wrong.  Your sub looks great!  There is a little star elongation--but no noise!

Rodd

Edited January 15, 2020 by Rodd

[Marvin Jenkins]

Thanks, for a single image I am very happy. I am using a Nikon D3100 on a SW 150pds without any extras. So next clear night I am going to try for a big batch of these and see what stacking can achieve. I promise I will frame it correctly though as the third galaxy is out of shot for no reason. See getting picky already after one frame.

 

Marvin

[Rodd]

2 minutes ago, Marvin Jenkins said:

Thanks, for a single image I am very happy. I am using a Nikon D3100 on a SW 150pds without any extras. So next clear night I am going to try for a big batch of these and see what stacking can achieve. I promise I will frame it correctly though as the third galaxy is out of shot for no reason. See getting picky already after one frame.

 

Marvin

You are well on your way--being critical of your work will improve your images.  If you are not guiding I would reduce the sub length to 60 sec and collect more subs.  This will keep your stars smaller and rounder, and if you stack 60 (an hours worth) you will be amazed.  If everything is working--let it run--an image made from 300 1 minute subs will be awesome. 

Rodd

[wornish]

Excellent image, I prefer the first.

How can anyone know what the most "natural image" looks like? - serious question what does that actually mean?

Does it not come down to personal preferences as to exactly how the image is post processed at this level?

If you like the image great if not then that's fine, we all are different.

Oh and we are all viewing the image on different screens/phones with different resolutions/ colour gamut etc.

Still I will say it again I like #1.

 

 

 

 

[Rodd]

51 minutes ago, wornish said:

Excellent image, I prefer the first.

How can anyone know what the most "natural image" looks like? - serious question what does that actually mean?

Does it not come down to personal preferences as to exactly how the image is post processed at this level?

If you like the image great if not then that's fine, we all are different.

Oh and we are all viewing the image on different screens/phones with different resolutions/ colour gamut etc.

Still I will say it again I like #1.

 

 

 

 

Thanks W.  I‘D say yes and no.....there are guidelines. I would not expect to see bright chartreuse Ha clusters in Olly’s image, or a smeary use of noise control by Barry W.    While an exact level of various attributes may not be specifiable, a general sense of “correctness” can be determined. No one would argue that a trees bark should be orange.  I like to think of Astrophotography as a kind of landscape photography.   For LRGB we are trying to depict space scapes Faithfully, as they would look if bright enough to see. The ocean is not beige (usually).  Most stars are not green or magenta.  The dust lanes of M31 have locations and dimensions.  A sense of depth is valued.  However, what you say about screens is oh so true.  Can we ever really know how our image look to others?  Sometimes I have trouble determining how mine look to me!

[ollypenrice]

I prefer the first as well. It's a very natural looking* image and the first is more relaxed. There are just a couple of things I'd look at if it were mine: the background sky is not a neutral dark grey but, without measuring it, my hunch is that green is too low. I'd also reduce its colour saturation (just the background, not the rest.)

Also the stellar blues are rather magenta which would square with the greens being a bit low.

However, I admire the hell out of this rendition.

Olly

*

10 hours ago, wornish said:

Excellent image, I prefer the first.

How can anyone know what the most "natural image" looks like? - serious question what does that actually mean?

Does it not come down to personal preferences as to exactly how the image is post processed at this level?

Good question to which there might be many answers but here are mine. 1) Many imagers have old-school film renditions in the backs of their minds when they look at digital images. These old images have a look which many of us think of as 'natural' though that is not a particularly rational use of the term. Cameras and film are not works of nature but, still, they influence our perception of 'natural-looking' pictures.  2) When you spend a lot of time processing images you develop an eye for the effects of particular processing inputs. Noise reduction, sharpening, colour saturation, star reduction, local contrast enhancement, stretching, setting of black and white point etc etc. You know what these actions produce if taken just a tad too far so a 'natural-looking' image is one in which they haven't been taken too far. They've been used to just below the point at which their signature becomes visible.  3) The imager's intentions and style: one person might want to extract every last scrap of information from an image through intensive processing. This is rather like trying to make a highly informative diagram and is like graphic design. Another might be trying to keep the processing invisible and accept that some details cannot be rendered without introducing a diagram-like look. Both are valid activities but the second aims to look 'natural.'

 

Edited January 16, 2020 by ollypenrice

[alan potts]

I personally like both of them, your work is always of the highest quality, M31 is not as easy to process as some may think, I have 5/6 batches of data from different nights on two or three different F/L's and Masking and stretching never seem to come out the same twice. As Long as I get a half decent picture I'm happy, just try to get a bit closer to the likes of you and Olly, maybe one day.

Alan

[Rodd]

2 hours ago, alan potts said:

I personally like both of them, your work is always of the highest quality, M31 is not as easy to process as some may think, I have 5/6 batches of data from different nights on two or three different F/L's and Masking and stretching never seem to come out the same twice. As Long as I get a half decent picture I'm happy, just try to get a bit closer to the likes of you and Olly, maybe one day.

Alan

Thaks Alan...i appreciate the kind words.  I agree about M31 not being easy.  The core is a slippery, demanding thing that seems to have a very strong chin (can't be knocked out!)

Rodd

[Rodd]

3 hours ago, ollypenrice said:

he background sky is not a neutral dark grey but, without measuring it, my hunch is that green is too low. I'd also reduce its colour saturation (just the background, not the rest.)

Thanks Olly--you have a keen eye indeed.  You are correct.  Also, the background is far to dark--I just looked at the numbers.  Can't do much about that, but I have equalized the colors and reduced them--so the background palette should be more neutral.  I have noticed some other things that I do not like--but fixing them would require a reprocess......thinking I had sumitted the mountain was wishful thinking...fooled by a false peak.  Anyway--here is the modified image

 

[alan potts]

I look at my brighter stars and thought for a long time they could be better and need a lot of work but I see mine now about the same as yours, notably the blue and the deep orange ones, so thats sort of pleasing for me, I don't get as much colour into the stars though. Probably because i don't collect enough data,. Two hours and I want to move on and see something else. I can see me going longer in collection when I have sort of done a year round the sky and then go back for more data. I don't dump any that is good so it keeps. Do like the core on these shots of yours you have managed a very nice translucency to it.

Alan

[Rodd]

3 minutes ago, alan potts said:

Two hours and I want to move on and see something else.

Same for me--I have to really struggle to not be a "kid in a candy store" on a clear night.  That's why I wanted to get the FSQ and .6x reducer so I could finish images in 1/4th the time--maybe even 1 night.  Didn't work out that way.  As far as stars--they are complicated.  My stars basically stink--cores white with colored edges.  There is a special tool in PI to repair stars but its complicated and I have not looked at it yet.  That is one thing that masked stretch does though--it doesn't overstretch the data which tends to keep the color in the stars.  It does create stars with tiny cores and larger fuzzy areas though--so I don't usually use it.  normally I try to make perfectly fitting star masks and work on the stars with curves (saturation) to try and give them evenly distributed color.  I did not spend much time on the star in this image.

Rodd

[alan potts]

I rather like the FSQ myself and would like one. As much as I can afford one, my wife could well have a use for one of my knives in the kitchen to do some filleting and I don't mean fish. Not in the same class but the Borg 77Ed ll impresses me with how fast it collects data for a reasonable image. The only thing is with using it with the 071 the field is wide and there is quite a vignette. Sadly I still have not got anything to deal with gadient.

Alan 

[Filroden]

I guess my only feedback is that I find the strong colours in the stars are distracting me from the galaxy. I'd be tempted to create a starless layer and only tweak the blue saturation in the galaxy disc. I'd also be tempted, in the star layer, to desaturate all stars a little (5-10% only) so that when recombined, the galaxy stands out a little more. Of course, this is totally unnatural, but the focus of the image should be on the galaxy and not the background! Had it been a photo, I'd have even put in an artificially vignette to further draw the eye into the centre of the image but it feels so wrong to do that having spent so long on flats to eliminate them in the first place

Or do nothing more to the last image and call it a good one!

[Rodd]

Just now, alan potts said:

I rather like the FSQ myself and would like one. As much as I can afford one, my wife could well have a use for one of my knives in the kitchen to do some filleting and I don't mean fish. Not in the same class but the Borg 77Ed ll impresses me with how fast it collects data for a reasonable image. The only thing is with using it with the 071 the field is wide and there is quite a vignette. Sadly I still have not got anything to deal with gadient.

Alan 

I have not seen a great improvement in data collection speed using the FSQ 106 at F3--still takes me multiple days to finish an image.   I like the scope for its wide FOV and quality components--but it did not live up to the "speed" hope that I had.  I believe that 2 things are needed for speed  1) a fast FR, and 2) aperture.  Is the FSQ at F3 faster than the FSQ at F5? maybe on paper--but it still takes me multiple days to finish an image.  The reducer may increase speed a bit by redistributing photons and pixels, but to really increase speed you need more photons--and a reducer can't provide that.  

[Rodd]

4 minutes ago, Filroden said:

I guess my only feedback is that I find the strong colours in the stars are distracting me from the galaxy. I'd be tempted to create a starless layer and only tweak the blue saturation in the galaxy disc. I'd also be tempted, in the star layer, to desaturate all stars a little (5-10% only) so that when recombined, the galaxy stands out a little more. Of course, this is totally unnatural, but the focus of the image should be on the galaxy and not the background! Had it been a photo, I'd have even put in an artificially vignette to further draw the eye into the centre of the image but it feels so wrong to do that having spent so long on flats to eliminate them in the first place

Or do nothing more to the last image and call it a good one!

Good call--I actually did those things (without going starless) in the last image--I could have desaturated the stars more I guess--but the stars can't be saved--They don't bother me too much.  I decrased their brightness a tad.