M31 colour adjustments.

[ollypenrice]

I've been convinced by vlaiv's arguments that galaxies are usually presented with too much colour and, in particular, with exaggerated blues in the spiral arms. I've already revisited my M101 and M33 so this time the Andromeda galaxy got the treatment. In one of his posts (which I can't find) vlaiv linked to the Hubble Team's rendition of part of M31. https://hubblesite.org/image/3476/gallery/73-phat

Using this for reference I found it wasn't difficult to get fairly close to it. Although the change from my original colour was striking, the adjustments were remarkably slight, raising the cyans and yellows in red to cool them down and tweaking the cyans and magentas themselves before lowering, considerably, their colour saturation.

Initially I wasn't too pleased with the result and found it lacking in punch. Waking up with fresh eyes I find I prefer it. I never liked the original colour anyway and have floundered around with it in the past. This version's growing on me. Edit: I also gave it the Starnet++ treatment since that wasn't available for the original.

 

(The original idea behind the image was to try to find the galaxy's outer limits. Main image FSQ106N/Atik 11000 mono with 30 minute luminance subs. Core, TEC140/Atik 11000. Mesu 200 mount.)

Olly

 

Edited November 15, 2020 by ollypenrice

[tomato]

I think there is some merit in this approach to the colour.

It’s interesting to me how the shape of M31 seems to change with how deep the image goes. It can either appear quite slim or as on one splendid example it can become more ‘compact’, resembling a giant version of NGC 7331. Here are a couple of my renditions to illustrate the point. Or it could just be an effect of the framing?

 

[Laurin Dave]

Very nice Olly..  us UK imagers can only dream of going so deep, it really does show the ‘twist’.  As for the ‘Outer Limits Galaxy’ I’d always thought that was nearby NGC891 😉. 
 

Dave

[ollypenrice]

47 minutes ago, tomato said:

I think there is some merit in this approach to the colour.

It’s interesting to me how the shape of M31 seems to change with how deep the image goes. It can either appear quite slim or as on one splendid example it can become more ‘compact’, resembling a giant version of NGC 7331. Here are a couple of my renditions to illustrate the point. Or it could just be an effect of the framing?

 

 

 

I think the framing has a lot to do with it. I'd like to add more background around this two panel, in fact, because I do like a galaxy to 'hang there in space.'  It's tempting to fit it on the diagonal but even full frame at 530mm won't get anywhere near to catching it all. I was originally motivated, on this project, to try to find what the star charts all show in terms of its outer reaches. The 30 minute L subs provided the breakthrough. I'm quite sure of that because I had a good set of 15 minute subs which weren't finding what I was hoping for.

Olly

[gorann]

Looks quite nice Olly and I also remember Vlaiv's @vlaiv suggestion that we tend to exaggerate the blue in galaxies. So you and Vlaiv now inspired me to tweak my 2.6 hours of RASA data on M31. The top image is the "original" and in the bottom one I have just brought down the blue and green curves a little bit. Made quite a difference. I have not slept on it yet so I have hard to judge which one I prefer🤔 The other question is if I should or should not collect Ha for the image. Adding Ha would certainly make it less true to what it really should look like to our eyes.

 

 

Edited November 15, 2020 by gorann

[ollypenrice]

10 minutes ago, gorann said:

Looks quite nice Olly and I also remember Vlaiv's @vlaiv suggestion that we tend to exaggerate the blue in galaxies. So you and Vlaiv now inspired me to tweak my 2.6 hours of RASA data on M31. The top image is the "original" and in the bottom one I have just brought down the blue and green curves a little bit. Made quite a difference. I have not slept on it yet so I have hard to judge which one I prefer🤔 The other question is if I should or should not collect Ha for the image. Adding Ha would certainly make it less true to what it really should look like to our eyes.

 

 

Second one for me, Goran. No doubt about it. Our 'outer glow' regions agree, I notice, but don't you want to do an extra panel on the left?  It's a shame to crop out something so rarely seen in images of M31 as that upward twist. Go on!!! 😁 (The outer glow on mine is somewhat over-stretched but that was a processing decision taken because the project had that in mind. Yours is more subtle. Maybe I'll ease mine down a tad at some stage.)

Olly

 

[Laurin Dave]

One of the things that struck me (apart from the clearly incorrect sensor spacing ) when I did this a couple of months ago was just how far out the Hii regions go - to the lower left almost to the edge of the frame (maybe even beyond), ie very close to the edge of the Halo in Olly's and Goran's images (it was outside what I could achieve in a reasonable time).   Here's the (superstretched) 13.5 hrs of Ha through a GT71/ASI1600 and the final LRGBHa image.

Dave

 

 

[gorann]

58 minutes ago, ollypenrice said:

Second one for me, Goran. No doubt about it. Our 'outer glow' regions agree, I notice, but don't you want to do an extra panel on the left?  It's a shame to crop out something so rarely seen in images of M31 as that upward twist. Go on!!! 😁 (The outer glow on mine is somewhat over-stretched but that was a processing decision taken because the project had that in mind. Yours is more subtle. Maybe I'll ease mine down a tad at some stage.)

Olly

 

I think I agree Olly (even if I have not slept on them), the second one is an improvement. But I bet that if we showed these images to the general public and let them have a vote, the first more colourful one would win.

[gorann]

40 minutes ago, Laurin Dave said:

One of the things that struck me (apart from the clearly incorrect sensor spacing ) when I did this a couple of months ago was just how far out the Hii regions go - to the lower left almost to the edge of the frame (maybe even beyond), ie very close to the edge of the Halo in Olly's and Goran's images (it was outside what I could achieve in a reasonable time).   Here's the (superstretched) 13.5 hrs of Ha through a GT71/ASI1600 and the final LRGBHa image.

Dave

 

 

Lovely M31 Dave! Maybe I should collect Ha for mine one day, maybe even do that before I make that extra left panel with the dust hook that @ollypenrice goes on about😉

[ollypenrice]

1 hour ago, Laurin Dave said:

One of the things that struck me (apart from the clearly incorrect sensor spacing ) when I did this a couple of months ago was just how far out the Hii regions go - to the lower left almost to the edge of the frame (maybe even beyond), ie very close to the edge of the Halo in Olly's and Goran's images (it was outside what I could achieve in a reasonable time).   Here's the (superstretched) 13.5 hrs of Ha through a GT71/ASI1600 and the final LRGBHa image.

Dave

 

 

 

Indeed the Hii regions really are widespread. I very much like your rendition, Dave. It has a softly softly look. But then that's part of the charm of this game: we don't just put out buckets, we seek to emphasize different features of an object and create a particular atmosphere.

Olly

Edited November 15, 2020 by ollypenrice

[billhinge]

Hi

If i may ask a question, how does any one know what the correct colour is? Surely it depends on the sensitivity of the sensor? CCD/CMOS/eyes? Isn't this similar to how do you know I see the same red apple as you? https://en.wikipedia.org/wiki/Aristotle's_theory_of_universals (we actually debated this at the ITU)

Surely you would need to measure the full spectrum of a body and come up with an algorithm that proportioned the colours according to wavelength intensity?

Also I find it interesting how people seem attracted to certain images, I like in your face detail but others seem to prefer pastel approaches, is either right or wrong if the detail exists in the real world?

[MartinB]

I have to confess when I have processed M31 I have made heavy differential blue adjustments to the outer arms.  I'm not sure how much I want my images to have true colour fidelity, drama and "prettiness" might come into the equation as well!  Just like day time photographs.

[Laurin Dave]

53 minutes ago, ollypenrice said:

Indeed the Hii regions really are widespread. I very much like your rendition, Dave. It has a softly softly look. But then that's part of the charm of this game: we don't just put out buckets, we seek to emphasize different features of an object and create a particular atmosphere.

Olly

Thank you Olly..   After completing  (?) M31 I was interested to know how its Hii regions compared to those in M33, particular NGC604 which is second in size in the local group to the Tarantula Nebula and which if it were at the distance of M42 cast shadows at night!   Here's a composite of M33 and M31 to scale, it made me think ... What the Milky Way would look like from outside  and  how lucky we are not to be near one!  also following this post maybe my M33 is too blue!   

Dave

ps apologies for the thread hijack 

Edited November 15, 2020 by Laurin Dave

[vlaiv]

54 minutes ago, billhinge said:

Hi

If i may ask a question, how does any one know what the correct colour is? Surely it depends on the sensitivity of the sensor? CCD/CMOS/eyes? Isn't this similar to how do you know I see the same red apple as you? https://en.wikipedia.org/wiki/Aristotle's_theory_of_universals (we actually debated this at the ITU)

Surely you would need to measure the full spectrum of a body and come up with an algorithm that proportioned the colours according to wavelength intensity?

Also I find it interesting how people seem attracted to certain images, I like in your face detail but others seem to prefer pastel approaches, is either right or wrong if the detail exists in the real world?

Because there is well established theory of color matching - colorimetry.

There are infinite number of spectra that produce same color in human vision - this is because human vision is trichromatic. It has three types of receptors and our brain "synthesizes" what we perceive as color based on level of stimuli from each one of those. Stimuli in each one of those depends on spectrum but in a way that includes "summation". If I asked you to name 3 numbers that add up to 12 - you'd be correct to respond with both 6+5+1 and 4+4+4. In the same way - two or three different spectra can produce the same color.

This has nothing to do with idea of you seeing "different sort of red apple" than I. It has to do with - if I show you two apples that I find looking the same - you'll find those two apples also looking the same. Hence the name - color matching.

In order to best understand what color matching is as opposed to just two people seeing the same image of something is as follows:

Imagine a scene that is imaged with two different cameras - say Canon and Nikon (or Sony, or insert your favorite brand here _____ ). After that, image from each camera is displayed on its own computer screen - one from Dell and another from LG (or again, maybe BenQ or insert your favorite monitor brand here ___).

Anyone who has seen the scene and sees image on computer screens, if all is well and cameras and computer monitors have been properly color calibrated, will agree that: scene had same colors as both first and second computer monitor.

This is definition of color matching - you are able to replicate color regardless of equipment used to record it or device used to display it and any observer agrees that color matches.

Things get a bit more complicated when we apply color to object as color that we see / record will depend on light source used to illuminate it, but in principle it is again down to "sum" of spectrum reaching our eye in the end.

For this reason, there is well defined color of astronomical object. In fact, because astronomical objects are illuminated with only one type of color - either one they emit themselves or in case of reflection nebulae or planets or similar - with nearby light source and we can't have different illumination conditions - things are even simpler than daytime photography where you need to be careful about type of lighting in order to recreate proper color of object (what is known as white balance).

Further, we know what color each type of star needs to be - physics tells us what sort of spectra we can expect from each star and colorimetry tells us how to convert that spectrum into color.

There you go - actual colors depending on spectral type, and if you want to be more precise, there are tables for stellar colors:

http://www.vendian.org/mncharity/dir3/starcolor/details.html

 

[ollypenrice]

54 minutes ago, Laurin Dave said:

Thank you Olly..   After completing  (?) M31 I was interested to know how its Hii regions compared to those in M33, particular NGC604 which is second in size in the local group to the Tarantula Nebula and which if it were at the distance of M42 cast shadows at night!   Here's a composite of M33 and M31 to scale, it made me think ... What the Milky Way would look like from outside  and  how lucky we are not to be near one!  also following this post maybe my M33 is too blue!   

Dave

ps apologies for the thread hijack 

 

No hijack, Dave. My M31 has appeared before and it's here again specifically to invite opinions on colour.

Olly

[Dezerker]

Hey Olly,

Possibly a very stupid question (I hope you can forgive me, I'm quite new to all of this!), but how have you kept your stars that size and brightness with 30 minute subs?  My limited experience of astrophotography so far suggests stars are doomed to become massive, chunky white blobs after even 5 or 10 minute subs, especially once stacked and processed, such that they drown out the surrounding space / pixels - is that an equipment quality thing, or a sky quality / moisture thing, or a processing skill?

A bright star is possibly a particularly difficult beast to tame, but I've seen M31 cores become blown out after subs way shorter than yours - do you have any tips, or possibly pointers to resources I can look into?

And I should have said - your image is incredible!  The detail!  And the colour of the stars...I would love to be able to take such images!

Edited November 15, 2020 by Dezerker

[ollypenrice]

1 hour ago, Dezerker said:

Hey Olly,

Possibly a very stupid question (I hope you can forgive me, I'm quite new to all of this!), but how have you kept your stars that size and brightness with 30 minute subs?  My limited experience of astrophotography so far suggests stars are doomed to become massive, chunky white blobs after even 5 or 10 minute subs, especially once stacked and processed, such that they drown out the surrounding space / pixels - is that an equipment quality thing, or a sky quality / moisture thing, or a processing skill?

A bright star is possibly a particularly difficult beast to tame, but I've seen M31 cores become blown out after subs way shorter than yours - do you have any tips, or possibly pointers to resources I can look into?

And I should have said - your image is incredible!  The detail!  And the colour of the stars...I would love to be able to take such images!

You're most kind!

The question you ask is a good one. Assuming you're in sharp focus you're in the lap of your optics, mostly, when it comes to capture and, while the scopes I used on this are excellent, they cannot give tiny stars on a hard stretch of long subs so it's mostly down to processing. If you're shooting LRGB you already have a set of effectively 'short' subs in your RGB. The point of luminance is to get more signal so, if you want a stack with less, you probably already have it in your RGB. Sometimes that's useful to bear in mind. 

On this image, made a few years ago, I used familiar star reduction techniques. An initial, gentle stretch was done with the stars masked, but this can only be done gently or it will soon show, especially where stars lie on faint nebulosity. It's a help but not the answer. After that I used the Astronomy Tools 'Make Stars Smaller' routine. This is great and was formerly known as Noel's Actions. Well worth having. However, the arrival of Starnet++ transformed star control in DS imaging. It's free and either Standalone or incorporated into Pixinsight.

Basically it removes the stars in a single click - but don't expect miracles. The output image often looks quite artificial and 'blotchy' where large stars have been removed. I took it into Ps and made a three layer stack with copies of the original top and bottom and the starless image in the middle. I wanted to replace the stars with smaller versions of the original so I set my top layer to blend mode lighten and, in curves, pulled down its brightness till the stars were tiny. In Blend mode Lighten they were now the only part of the original showing in the blend. By playing with the curve I could get them to look crisp, natural and small. Trial end error. Flatten top onto middle when happy. In one or two places (the satellite galaxies) Starnet had done some damage but I had an original as my bottom layer so I could just erase the remaining top layer where necessary.

One or two stars needed a cosmetic fix post starnet but nothing drastic.

Olly

Edited November 15, 2020 by ollypenrice
Typo

[andrew s]

I have a copy of James D Wray's "The Color Atlas of Galaxies ". Unfortunately,  it does not include M31. It does however, very carefully ensure that the images are accurate on the UBV photometric system. While this if different from RGB it is interestin that most of the images are considerably more muted than typically presented images. 

I am sure they would be considered too dull if posted nowadays. 

I will try to get an image to post taken in daylight as trying with led .ight fails to reproduce the colour balance.

Regards Andrew 

[vlaiv]

11 minutes ago, andrew s said:

I have a copy of James D Wray's "The Color Atlas of Galaxies ". Unfortunately,  it does not include M31. It does however, very carefully ensure that the images are accurate on the UBV photometric system. While this if different from RGB it is interestin that most of the images are considerably more muted than typically presented images. 

I am sure they would be considered too dull if posted nowadays. 

I will try to get an image to post taken in daylight as trying with led .ight fails to reproduce the colour balance.

Regards Andrew 

I always wondered what is exact definition of for example UBVRI system?

If one does search on google for UBVRI response curves - they can hardly find two graphs that look the same - and these are just filter response curves and not combined instrument + filter response curves.

How does one account for instrument + filter response to obtain same results as someone else using different instrument and filter set?

For example:

vs

vs

vs

[andrew s]

@vlaiv they were once well defined. I will see if I can find the original spec. It certainly involved a particular photomultiplier tube and if I remember correctly involved it's cutoff in U! 

Regards Andrew 

Just checked V was cutoff at long wavelength by the PM tube and U by the atmosphere.

Still looking for an accessible reference but it looks like issue of consistency stated  very early after it's introduction in the 1950s

Edited November 15, 2020 by andrew s

[Mr Spock]

I think quite a few images have too much colour. But, how do we know what's right and what isn't? At some point the end result will be subjective and depend on personal taste.

[vlaiv]

4 minutes ago, Mr Spock said:

I think quite a few images have too much colour. But, how do we know what's right and what isn't? At some point the end result will be subjective and depend on personal taste.

It's rather simple - we know what is right because we measure things.

As long as you stick to measurements and standards - you'll get objective thing. As soon as you start doing things the other way - it becomes subjective.

There is well a defined color standard / absolute color space - CieXYZ. There is well defined relative color space that we all use on internet - sRGB. There is well known transformation between the two. Only thing that is missing is to know transformation between your particular camera and these color spaces (you can actually go directly from camera to linear sRGB bypassing XYZ if you want since all transforms are linear).

When you choose certain white balance preset in DSLR - you apply one of these transforms that converts what has been captured into actual color. These have been preloaded at factory.

For astronomy purposes, we don't really need to do that much work as DSLR makers do since they need to account for different lighting conditions (cloudy, sunny, inside, tungsten, halogen ...). In astronomy, colors don't depend on ambient light - they are light and we capture it.

We just need proper transform from our linear data into linear sRGB and we need to apply proper gamma function for sRGB color space.

 

[andrew s]

@vlaiv  this provides a good review . I will still try to find the original filter specs for amusement.

Bessel2005ARAA43p293.pdf

Regards Andrew 

PS see table1 in this

low_cost_BVRI_filters.pdf

Edited November 15, 2020 by andrew s
Added paper

[andrew s]

Here is an image of M51 from Wray's book

Regards Andrew 

[vlaiv]

2 minutes ago, andrew s said:

Here is an image of M51 from Wray's book

Here is HST image for comparison:

and "stellar bar".

I think that HST images in general match star colors very well - which leads me to believe they have been properly calibrated for display.