Keep close to authentic look of object or create stunning, award winning and over processed photo?

[sn2006gy]

What I'd like to see is continue to support the duality of science/art but allow people to contribute their subs/unprocessed masters as part of a greater collection where the science of consumer astrophotography is gathering more data than any other way would be feasible. By this, i mean allow us to share our masters and use science to improve the integration, calibration and processing techniques thereof.   You know, move to massively parallel calibration and GPU assisted integration & stacking.  This way we could be creative with what we capture locally but share the fruits of our labor and help improve data on a global scale.  Could be kind of cool. 

Most professional observatories are already moving to parallel mirrors and collection across multiple scopes in lieu of bigger and more cumbersome scopes.  Imagine if we helped usher that in for amateur astrophotography

[ollypenrice]

On 01/01/2021 at 00:14, saac said:

 

Now ask me what I think of Tracey Emin's bed  or the formaldehyde shark!

https://www.tate.org.uk/art/artworks/emin-my-bed-l03662

 

Jim 

As an occasional writer of fiction (for fun) I penned a short story lampooning 'conceptual art.'  A pretentious critic with a wealthy mother is looking at the Emin bed and muses, 'Although beyond his own means it would probably not be beyond those of his mother but, when he had once dropped a hint, she had replied that she had beds of her own and someone to keep them tidy.'

On 31/12/2020 at 23:43, vlaiv said:

Would you characterize following scenario as "authentic" or "artistic":

You take DSLR camera, set it to faithful color reproduction and take an image of a scene and transfer it to computer without any further processing.

 

I suspect that, for the purposes of the OP's question and this imaging debate in general, the answer would be 'yes.'  Different systems will produce different results but not radically different ones, at least not radical enough to offend what most people mean by faithful colour.

Consistency: a few years ago I posted up a Christmas dataset on here for folks to process as they wished. It was HaLRGB on the Cave Nebula, so a broadband target enhanced by Ha. Barry Wilson processed a rendition entirely in Pixinsight. My own was done in Photoshop. They were, to my eye, very much the same and, if forced to identify my own against Barry's in a blind test after a week, I'd have struggled. When Tom O'Donoghue and I decided to go after the IFN in our joint dual rig we were interested in the colour of this nebulosity. At that time it was usually seen as fairly grey but we threw a lot of RGB time at it and deliberately processed it independently. Our results were nearly identical. We found it to be brown, as do others who've gone after it.  If we exclude colour mapped images are LRGB/OSC images really as varied as some suggest?

Olly

Edited January 2, 2021 by ollypenrice

[vlaiv]

1 hour ago, ollypenrice said:

I suspect that, for the purposes of the OP's question and this imaging debate in general, the answer would be 'yes.'  Different systems will produce different results but not radically different ones, at least not radical enough to offend what most people mean by faithful colour.

Well, for all those interested in practical answer to this question, here is interesting website:

https://www.dpreview.com/reviews/image-comparison/fullscreen

And here is screen shot of that tool:

You can select up to 4 different cameras to be compared in various parts of above test panel and look at results.

Just to be further clear with what I'm trying to say:

For me, camera is measurement device. Expecting two different cameras to produce different results and being fine with that is like expecting two rulers to measure different dimensions of the box and saying - you know, it depends on who is measuring and there is artistic freedom in how you measure

Artistic freedom comes after you measure - in the way you write your numbers down. You can scribble them on a napkin or you can do some calligraphy wonder carved on a piece of frosted glass. As long as you keep the numbers same - so everyone can tell dimensions of the box - you preserve authenticity of measurement.

Here we are discussing something else - what happens when you can no longer read the numbers properly. That is the part I'm concerned with. To further extend this analogy, not doing proper color calibration and not being careful what color space you are using in the end and doing proper transform to that color space - is like measuring in some arbitrary units of length (like 1/3 of an inch is basic unit), writing numbers down and then not telling anyone what units you used to measure the box.

Take two such results - one written on napkin and one carved on frosted glass and there will be difference in artistic impression but also - if you compare numbers, you'll find them different and wonder - how on earth could these two represent same box?

1 hour ago, ollypenrice said:

When Tom O'Donoghue and I decided to go after the IFN in our joint dual rig we were interested in the colour of this nebulosity. At that time it was usually seen as fairly grey but we threw a lot of RGB time at it and deliberately processed it independently. Our results were nearly identical. We found it to be brown, as do others who've gone after it.  If we exclude colour mapped images are LRGB/OSC images really as varied as some suggest?

I'm still not convinced that IFN and dark nebulae have reddish/brown tone to them, and I'll explain why.

First - here is sort of a quiz question for everyone. Say you have jpeg image and it has color (240, 48, 31) (240, 48, 21):

If you were to take that color on your screen and cover it with ND filter that removes exactly 1/3 of light and make another square with different RGB values - what RGB values should you use in order to match color between filtered and this squares?

Answer to this question has to do with brown color of nebulosity among other things.

Second thing that causes me not to accept brown color as color of interstellar dust (I was split second away from making a joke on dark matter here) is this:

Yes, these are image of the same object, and yes - one is white-yellow while other is yellow-orange. Both are shot with presence of something that obviously alters color of the object - a sort of filter that seems to remove blue parts of the spectrum when you shoot thru it. Yes, atmosphere

Look at this graph:

This is in magnitudes - which means that is even steeper in ratios.

So what do we get when we apply that model?

If we start with nice gray:

and we reduce green some and blue some more:

What, brown? No way!

 

Edited January 2, 2021 by vlaiv
typo

[andrew s]

You can and and spectroscopists and photometrists do correct for atmospheric absorption.  It is strongly dependant on zenith angle. Good enough models are available which could be used at a subs level to correct a stack taken over an extended period. 

However, is it not normally taken into account when colour balancing the stars in an image? I don't know.

Regards Andrew 

[andrew s]

Here is a plot of atmospheric absorption from 4000 to 7000 A linear scale 0 to 1 at 45 deg and 90 deg from the horizon. It uses a typical AOD (aerosol optical depth of  0.13 for  France.)

Regards Andrew

Edited January 2, 2021 by andrew s

[vlaiv]

10 minutes ago, andrew s said:

You can and and spectroscopists and photometrists do correct for atmospheric absorption.  It is strongly dependant on zenith angle. Good enough models are available which could be used at a subs level to correct a stack taken over an extended period. 

However, is it not normally taken into account when colour balancing the stars in an image? I don't know.

Regards Andrew 

When doing a stack, if you select a reference frame and do linear fit on that to normalize frames - then I think you don't need to do any special processing for changing over imaging session - as each other sub will be normalized to exact time of selected reference frame.

After stacking it is rather easy to do color correction for atmosphere - you select couple of stars and calculate correction based on their stellar class / color index (similar to differential photometry). Only thing you need to be careful about is interstellar reddening.

However - people don't do above - in part as it is not available as simple "click here" option in processing software and doing it by hand requires both knowledge and tools.

Most popular software package for processing PixInsight - has tool that I expected to do above - but it turned out that it does not. It is named Photometric color calibration - and that name suggests above process, however it does not do actual color calibration in terms of image - it does some sort of "color index" calibration for photometry.

You can read details here:

https://pixinsight.com/tutorials/PCC/

 

[vlaiv]

1 minute ago, andrew s said:

Here is a plot of atmospheric absorption from 4000 to 7000 A scaled 0 to 1 at 45 deg and 90 deg from the horizon. It uses a typical AOD (aerosol optical depth of  0.13 for  France.)

Is that in magnitudes or? I don't see any units on the graph

[andrew s]

5 minutes ago, vlaiv said:

Is that in magnitudes or? I don't see any units on the graph

Linear units. I was just about to add that! The atmosphere cut off very steeply below 3900 A Regards Andrew

 

PS Theory here http://www.astrosurf.com/buil/extinction/calcul.htm implenented in his ISIS software.

Edited January 2, 2021 by andrew s

[ollypenrice]

3 hours ago, vlaiv said:

 

 

What, brown? No way!

 

'As for the color, obviously I didn't have enough data to get an accurate rendition of the color in the IFN, and what I've got wasn't deep at all. I knew beforehand I wouldn't have enough time to get deep and detailed color data, so I compromised with at least getting enough color for the stars and the field, hoping that the IFN would at least inherit some color from the background signal, which is most definitely the case here, and that's why the IFN has a brownish hue versus the more expected blueish cast - regardless, the IFN not only scatters blue light but is also fluorescing a broad red spectrum of light known as the Extended Red Emission (ERE), so the brownish hue acquired from my poor color data isn't completely off track.'

That's from RBA's website: http://www.deepskycolors.com/archivo/2010/04/08/integrated-Flux-Nebula-Ifn-really-wide.html

For what it's worth, my own IFN: 

Same as Rogelio's. So what's your conclusion? 'It isn't brown, it only looks brown?' 

Olly

[Martin Meredith]

https://clarkvision.com/articles/color.of.nebulae.and.interstellar.dust/ is quite interesting (it aint brown enough!).

Also this: https://clarkvision.com/articles/astrophotography-color-and-critics/

Martin

[mftoet]

In the end: brown is dark orange. It’s also called a ‘context colour’. Here’s a YouTube video about brown not being an actual colour.
 

 

[vlaiv]

24 minutes ago, ollypenrice said:

Same as Rogelio's. So what's your conclusion? 'It isn't brown, it only looks brown?' 

My point is that actual color of object is not the same color that you have in your images.

I listed two possible reasons and I believe that brown color in most images is due to combination of the two.

One in a form of a quiz question - and that one is particularly interesting as it will show something many people do not expect - and is very much related to the whole color thing.

Other is the fact that our atmosphere shifts color of object to red by scattering shorter wavelengths more than longer (for that same reason sky appears blue and not white).

So if you take what is inherently grey object and observe it thru our atmosphere it can easily happen that it looks brown instead - and you conclude that object is brown - while it is grey in reality.

13 minutes ago, Martin Meredith said:

https://clarkvision.com/articles/color.of.nebulae.and.interstellar.dust/ is quite interesting (it aint brown enough!).

Also this: https://clarkvision.com/articles/astrophotography-color-and-critics/

Martin

There are many things where I don't agree with Clark. In fact, many statements can be easily shown to be incorrect.

Take for example this:

This is from the page on nebula color. This is XY chromaticity diagram (incorrectly displayed on sRGB profile while showing colors of whole human vision gamut). Simple math shows that colors that are generated as linear combinations of sources (points on this diagram) - lie on lines connecting these dots.

Above marked area is much smaller than what can be generated by marked narrow band sources. In fact, proper diagram explaining that would be:

Any color inside marked outline can be generated by source that consist from linear combination of marked single wavelength sources (hydrogen alpha - gamma, OIII and helium I).

If we want to be more accurate - we will actually use proper XY chromaticity diagram that can be show in .jpeg/png image, and image would look like this:

This really means that these sources are capable of generating almost all colors our display devices (computers and phone screens that support sRGB profile) are capable of showing except for saturated greens which lie outside of marked region.

Btw, notice that sRGB triangle is actually - triangle. This means that any sRGB color can be produced with Green, Red and Blue components in linear mix - and actual Green, Blue and Red - have coordinates of vertices in triangle - again showing that any linear combination will lie on line / region enclosed by set of points that have chromaticity of sources.

[mftoet]

I have a lot of images showing brown(ish) dark nebula. I don’t specifically try to bring out this colour or hue, it’s simply there. I always try to white balance my images on a the background sky (which can be hard to find when there’s lots of nebulosity in the frame. Furthermore I often increase the vibrance of the colours in my images by 20 to 30 in the Photoshop RAW converter.

The last example is a rather interesting one: apart from what might be a gradient in the lower left corner, different hues can be distinguished in the molecular clouds: ‘50 shades of ...’?

Edited January 2, 2021 by mftoet

[Martin Meredith]

If you apply your (non-contentious) statement that our atmosphere reddens ('de-blues') by scattering, is it so hard to accept that the same scattering of shorter wavelengths by dust beyond our atmosphere leads to reddening (browning, and not greying) in Olly's image?

Martin

[andrew s]

I tried a very simple experiment. I created a grey rectangle in powerpoint. I then reduces the green and blue by the factors calculated from the atmospheric absorption from the 45 deg plot I posted. Here is the result.

While I should have integrated over the filters response, I just used the centre wavelengths. However, to first order it's obvious the atmosphere will make grey look brown if you don't compensate for it one way or another.

Regards Andrew

Edited January 2, 2021 by andrew s

[vlaiv]

13 minutes ago, Martin Meredith said:

If you apply your (non-contentious) statement that our atmosphere reddens ('de-blues') by scattering, is it so hard to accept that the same scattering of shorter wavelengths by dust beyond our atmosphere leads to reddening (browning, and not greying) in Olly's image?

Martin

It's not hard at all to accept anything that we can verify with measurement. On the other hand - I know how people often process their images and I know what in that process might lead to nebula showing brown color when in fact it is not.

There is rather simple way of finding out what the actual color of dark nebula is.

- Take set of color calibration images and devise color transform matrix for your setup (that will produce true color in the sense we discussed above - or if you will, if you shot the scene I linked above and compared to any other camera -  you would get the same colors as all other cameras - within gamut differences)

- Do photometric measurement on calibrated data

- Devise linear R, G and B values (or X, Y, Z - any color space really) and represent that color in sRGB color space to be shown here on SGL as color sample (an thus be accessible to most people).

 

[vlaiv]

7 minutes ago, andrew s said:

I tried a very simple experiment.

Did you account for gamma function of sRGB color space ?

That is another factor to consider.

[andrew s]

48 minutes ago, vlaiv said:

Did you account for gamma function of sRGB color space ?

That is another factor to consider.

No. I just multiplied the 128 by the ratio of the absorption. I did say it was simple.

Regards Andrew 

PS I just use a mono camera to take spectra no fancy colour processing. 

Edited January 2, 2021 by andrew s

[andrew s]

56 minutes ago, Martin Meredith said:

If you apply your (non-contentious) statement that our atmosphere reddens ('de-blues') by scattering, is it so hard to accept that the same scattering of shorter wavelengths by dust beyond our atmosphere leads to reddening (browning, and not greying) in Olly's image?

Martin

Interstellar reddening is very real and well known.

Can someone send @vlaiv a set of data to calibrate in the way he proposes then we will know.

Certainly, the IFN will redden any light passing through it. 

Regards Andrew

Edited January 2, 2021 by andrew s

[vlaiv]

1 minute ago, andrew s said:

Certainly, the IFN will redden any light passing through it. 

Will it redden also surrounding light or will that be blue shifted?

In fact, what will be the primary source of light coming from IFN? Background lighting or surrounding start light?

[andrew s]

1 minute ago, vlaiv said:

Will it redden also surrounding light or will that be blue shifted?

In fact, what will be the primary source of light coming from IFN? Background lighting or surrounding start light?

I think it will depend on the detailed circumstances.  You get blue reflection nebulae with back scatter. I don't know what the primary source of IFN is. True colour photometry might help decide what it is. Back or forward scatter, or absorption of transmitted light. 

If the surrounding light does not pass through it then it should have no effect. It would require something to scatter additional light into our line of sight to change it. 

 

Regards Andrew.

[ollypenrice]

I'm sorry but I have never pretended that my images were taken from anywhere else but planet Earth  and my understanding of 'natural colour' is 'colour as seen from here.' What are we to make of, say, IC342 which is not only reddened by the earth's atmosphere but by intervening galactic dust? In the link I've compared it with M101 for size but it can also be compared for colour, as can the stars and background. I would be very surprised if, intervening dust permitting, there were much difference between the two galaxies but what I'm trying to do is photograph them from here.

https://www.astrobin.com/full/327910/0/

Olly

Edited January 2, 2021 by ollypenrice

[vlaiv]

3 minutes ago, ollypenrice said:

I'm sorry but I have never pretened that my images were taken from anywhere else but plaet Earth and my understanding of 'natural colour' is colour 

Just for the sake of argument, can you respond to that quiz question?

[ollypenrice]

3 minutes ago, vlaiv said:

Just for the sake of argument, can you respond to that quiz question?

😁Tomorrow! I've had rather a long day!! 

Olly

[vlaiv]

7 minutes ago, ollypenrice said:

😁Tomorrow! I've had rather a long day!! 

Olly

Sure.

I know that you are not pretending to shoot from somewhere else, but sometimes you remove influence of atmosphere without giving it much thought. For example - that IFN is at least 5 magnitudes fainter (so IFN is 1% or less of sky signal) than natural sky and you can't show it unless you remove sky glow in the image.