33 replies to this topic

[maxsid]

There are thousands of images of M31 on the internet but the colors are all over the place.

Is there a reference image (not one of those artificially-looking-hubble-like-narrow-band)?

Or it's purely subjective?

 

Recently, I posted this image https://www.cloudyni...2edii-11-hours/

The community here and my friends/family liked it.

 

 

Today I added two more hours of exposure from a darker site and I got this. With about the same processing except for different star color correction in APP and star reduction in Photoshop.

 

 

Which one is truer?

Edited by maxsid, 01 December 2020 - 04:11 AM.

[jerr]

In fact the only true color is grey.

The rest is our imagination and liking.

And as it comes to liking the bottom one looks really neat!

Edited by jerr, 01 December 2020 - 12:00 PM.

[Sandy Swede]

To pick up on what jerr said (and I agree with jerr's view), here is a link to a great article in Nautilus concerning the relationship between color and perception:

http://wise.nautil.u..._eid=a312f0b998

 

My vote is for image #2.  Perhaps it is just aesthetics, but I like the blue in the rim area.

[Tapio]

If it's photometric color calibrated then that is good enough for me.

Both look good, latter maybe better.

[lee14]

No, the true color of M31 is not 'grey'. Luminous objects appear gray to the eye when the light levels are too low to stimulate the cone cells of the retina. Longer photographic exposures can compensate for this. There is physical reddening resulting from selective absorption by dust, blue from reflection nebulae and other factors, and brighter reds from H alpha regions. Both OP images are very good, but the second is excellent!

 

Lee

Edited by lee14, 01 December 2020 - 08:14 AM.

[WoodlandsAstronomer]

Interesting post OP. You certainly see some creative works out there that admittedly look more like an artists rendering than a photograph. I’ll take a stab at an answer to your question...

The true color of light corresponds to the wavelength of the emitted light. Different types of ionized gases have their telltale signature wavelengths and color, hydrogen alpha, oxygen III, sulphur II, etc. so theoretically, if you had a uniformly sensitive color sensor (does such a thing exist? Trick question) you would arguably be capturing the true color in the right amounts of each. However, there are several hiccups, the primary being that color sensors, or more to the point Bayer matrices on top of greyscale sensors, are NOT uniformly sensitive and tend to favor one color more than others. Another issue that may change the received color versus the transmitted color is red/blue shifting that shifts the wavelength of light we receive one way or the other depending on whether the object is moving away or towards us. Another issue is that deep space objects that are exceptionally far away and faint, require us to go deeper and longer to, for example, bring out the reddish nebulously in the spiral arms of Andromeda.

Narrowband and LRGB imagers address these shortcomings by isolating wavelengths or band passes to maximize data and then the imager constructs a final image balancing the contribution of various wavelengths to the final image and sometimes mapping it to, for example, the Hubble palette. These tend to show much more detail and are of more scientific interest, but faint hydrogen alpha may be leveled up to create a more pleasing image that may not correspond to comparative levels in relation to oxygen III, for example. Sometimes these images get a little too artistic looking for my taste than photographs, depending on the skill or taste of the processor of course.

If you could get over the color matrix issues of OSC and capture color uniformly, this would theoretically present the true balance of colors and likely be more representative of the true color of an object.

Edited by WoodlandsAstronomer, 01 December 2020 - 07:43 AM.

[RareBird]

I read that recent improvements to PixInsight promise to greatly improve photometric color calibration by using the Gaia star catalog as a reference. By knowing the amount by which each star’s color must be corrected in an image, you can address not only color calibration, but vignetting and gradients. That’s what should be ‘coming soon’ to PixInsight, if I understand correctly. As promising as that sounds, I think the calibrated results will leave most users wanting more color and dialing up that saturation to +11, anyway.

[DRK73]

I'm really don't think that the OP is asking some deeply speculative and philosophical question about the nature of light, the physics of light, blackbody radiation, how eyeballs work, emission spectra, and all that other jazz...

 

I *think* OP is simply asking he (?) got the color about right on his image of M31. 

[idclimber]

In my mind one of huge draws to AP is displaying things your eyes could never see. If you really want the images to look natural you can always go back to an eyepiece and see a smudge. 

 

Much of the post processing we do to our images is subjective and a matter of taste. 

[ssa2294]

The first photo looks almost exactly like the first ones I took, except I did not have that many subs. Then the 2nd photo looks more like what you can expect with longer exposure (=more subs). Just for giggles I ended up combining almost all my nights of M31 after processing at the same time combining them all. You do see with longer and longer integrations the change into blue. What I am still missing and unable to get are the red regions. 

 

Ultimately what maxsid is asking is same question I have asked myself, if M31 for example was to drive right up to my front door and park and I was to take a photo of it as proof...what colors would I see? When using PixInsight I have at times nudged the saturation too much, or something else too little to end up with an M31 that would be fit logo for either Minnesota Vikings or Baltimore Ravens.  It is a good question to ask, wish I had an accurate answer. What is the true color or most accurate representation we can find, so when we do process we don't have too much blue or too little yellow. 

 

So, short of building my own starship with faster than light capabilities, who does have the most accurate photos of these whacky objects we are shooting?

[Sandy Swede]

I'm really don't think that the OP is asking some deeply speculative and philosophical question about the nature of light, the physics of light, blackbody radiation, how eyeballs work, emission spectra, and all that other jazz...

 

I *think* OP is simply asking he (?) got the color about right on his image of M31. 

Party pooper!  We'll get you for that -  and your little dog, too!   

You ask a bunch of folks with years of background in astronomy, many of whom have degrees in the sciences, a question like this and you are surprised at receiving somewhat arcane or techie responses?

[maxsid]

Jerr,

I can't agree with this statement.

Maybe grey for the human eye but not to a telescope/camera.

There's a lot of color in the images, don't you agree?

 

In fact the only true color is grey.

 

[maxsid]

Yes, pretty much this.

I guess if I saw a properly photometrically calibrated image that would give me the answer.

 

I *think* OP is simply asking he (?) got the color about right on his image of M31. 

[fewayne]

There are thousands of images of M31 on the internet but the colors are all over the place.

Is there a reference image (not one of those artificially-looking-hubble-like-narrow-band)?

Or it's purely subjective? [...] With about the same processing except for different star color correction in APP and star reduction in Photoshop.

So maxsid appears to be actually going for photometry and blackbody and all the rest of it. No little dogs need be harmed in the making of this post.

 

Maxsid, what did your star calibration graph look like in APP? I mean, even the interpretation of fit for those probably leaves something to the esthetic eye, but if you set the kappa to eliminate the outliers, lanced the trendline through the main body, and set the white crosshairs at something like 30% of the climb up the population...unless Mabula argues with us, I think that's going to be pretty accurate.

 

You did use "Adaptive Black Body & Extinction" for your model -- RIGHT??! I wouldn't want to think you're one of those ARTSY types that just makes an image LOOK good.

[audioengr]

The center is off-white, slightly orange, the next bands are reddish and the outer bands are blue.  You have to stretch the intensity with curves and saturate to see colors, particularly the blue.  My colors are based on a black-body model that knows that 80% of the stars are reddish and 20% are blueish, so I'm pretty sure they are close. It analyzes thousands of stars and maps these to the colors.

 

[maxsid]

Actually, no... The difference is not because of more subs (in my case). Just a bit different star color calibration in APP.

BTW, APP doesn't use photometry, it uses Adaptive Black Body and Extinction statistics.

 

Here's two images. Same data-set for both (same as for the first image in this post).

Processed only in APP, no Photoshop.

-- stretch (15%, 3 sigma, 2.5% base)

-- saturation (default)

-- star color calibration (a bit different for each image)

-- no further tweaking.

 

image 1 - the whole image was used for calibration

 

image 2 - six boxes (avoiding the galaxy) were used for calibration

 

 

The first photo looks almost exactly like the first ones I took, except I did not have that many subs. Then the 2nd photo looks more like what you can expect with longer exposure (=more subs). Just for giggles I ended up combining almost all my nights of M31 after processing at the same time combining them all. You do see with longer and longer integrations the change into blue. What I am still missing and unable to get are the red regions. 

Edited by maxsid, 01 December 2020 - 08:41 PM.

[WadeH237]

Spiral galaxies tend to have older and cooler stars in their cores, with younger and hotter stars in the spiral arms.

 

In general, this means that the cores should be somewhat yellow and the arms should be somewhat blue.  There are likely going to be star forming regions scattered about and they will tend to be the red color of Ha emissions.

 

One of the ironies in beginning astro imaging, is that some of the brightest objects, like the Andromeda Galaxy and Orion Nebula for example, are not some of the easiest to process.  The issue with Andromeda is that the core is very, very bright and and there are extents that are very, very faint (and they extend further than any of the images here).  Most of the images that we see show mostly the core and the main arms.  The bluest parts are really faint and don't show up in most casual images, so we're used to looking at yellowish Andromeda images.  Sometimes people try and coerce the blues out by shifting the color balance.  The issue with that, it makes the entire image too blue.  To get the blues out of the spiral arms, you need to go pretty deep and you need to do non-linear saturation adjustments to make those faint areas blue without shifting the colors of the rest of the image.

[maxsid]

I think I got my answers here.

Thank you very much guys!

[jerr]

Jerr,

I can't agree with this statement.

Maybe grey for the human eye but not to a telescope/camera.

There's a lot of color in the images, don't you agree?

Max,

 

That's the point. 

Are beauty pictures by Chandra having any (true) color?

[erictheastrojunkie]

Max,

 

That's the point. 

Are beauty pictures by Chandra having any (true) color?

Save the philosophy for some other forum, there is absolutely a science behind color AND perception of it, and no, grey is not the only true color. (and fwiw, which isn't a lot, I did my graduate education on protein expression in retinal endothelial cells and how it relates to disease development and progression within the eye...affecting color and light perception). 

[jerr]

Sure thing Eric.

[bobzeq25]

Color in our images is _unavoidably_ partly subjective.  Beware the "my color is more realistic than your color" discussions, they're pointless.

 

Real astronomers do not use color, they gather quantitative data.  Numbers.  If you see a "professional" color image, it was done for public outreach.

[dhaval]

Colors!!! Even within our own species, not everyone has the same perception of color, much less with other living organisms. And who knows, if aliens exist, what they may think of color. 

 

This really is one of those popcorn discussions.  

 

CS!

[idclimber]

Colors!!! Even within our own species, not everyone has the same perception of color, much less with other living organisms. And who knows, if aliens exist, what they may think of color. 

 

This really is one of those popcorn discussions.  

 

CS!

And even if you processed the image "perfectly" with "scientifically validated color" on a perfectly calibrated monitor it is still posted on the internet and shared with our friends where where most of them view the image on a monitor that might as well been calibrated by a monkey. 

[sn2006gy]

Color isn’t subjective. How one changes color to create a more pleasing image is.

There are more precise ways to measure temperature and brightness and distance and other scientific persists however color does reflect what we know about stars such as young blue stars or old orange stars or medium life average yellow and white stars

Take Pleiades for example. A good color image will show the reflection of the bright blue stars across the vast dust and said color image would also show the dust fade to gray further away from the central blue stars.

Our eyes have capability to see colors that may different between people/species but that doesn’t mean the color isn’t real. We just see in what we refer to as the visible spectrum.

We have technology that allows us to see in a much broader range of the spectrum..

None of which removes what color is.

However.. remember color is relative. Not because of subjectivity but because of interstellar space/dust/time/distance - seeing andromeda as it was 2.5 million years ago through the noise of our own galaxy, dust, gas, atmosphere, light pollution and choice of color filters or color camera and it’s dynamic range so on and so forth - we just get the color because our our cosmic relationship.

Picking up less sodium photons may allow your sensor to grab more blue photons