439 replies to this topic

[loujost]

Welcome to CloudyNights, mate!

 

The truth is, I can't even get 24 colors (technically, 18 colors and 6 grays) to match with a single least-squares-fit CCM. Tossing in more colors will just create a one-size-fits less matrix. Only a lookup table can compensate for the metameric failure of the camera.

 

And to what end? If I'm shooting a particular object with a known measurable "truth", I don't need to optimize across the entire spectrum. I can just generate an optimized CCM for the chroma zone of interest. But even then, the exercise would be entirely academic: it wouldn't fix the metameric mismatch between even normal retinas:

 

"The images shown in Figure 3 illustrate the cone mosaics for five human subjects. In each image, the L cones are indicated in red, the M cones are in green, and the S cones are in blue. Each image was obtained from a different subject, and all subjects had normal trichromatic color vision according to standard tests. Several striking features of the mosaics are apparent."

[Source: D. Brainard, "Color and the Cone Mosaic", Annu. Rev. Vis. Sci. 2015. 1:519–46]

 

 

So I have no idea if what I see as "red" is the same as what you see as "red". Because individual perception (of any sense) is random, the 1931 CIE color space is based on the measurements of 10 (or 7) standard observers. It's much like the Scoville Scale, where 4 out of 5 people must agree that this chilli is red hot.

 

"A weakness of the Scoville organoleptic test is its imprecision due to human subjectivity, depending on the taster's palate and number of mouth heat receptors, which vary widely among subjects. Another shortcoming is sensory fatigue; the palate is quickly desensitized to capsaicinoids after tasting a few samples within a short time period. Results vary widely (up to ± 50%) between laboratories." [Source]

 

BQ

As we mentioned m earlier discussions, some kinds of variations across individuals have no effect on color matching. The question is not "Is his experience of red the same as mine?" (a very difficult question to answer) but rather "Does his experience of of the image match his own experience of the object?"

[Robert7980]

When in doubt the photo-spectrometer doesn’t lie… Not difficult to precisely measure the “color” as a function of energy content… If colors weren’t understood to ultra high precision almost none of our technology would function. Suprised there is such debate on something that’s so exquisitely defined and measured without ambiguity in a ways that are repeatable and predictable. Not sure there was as much content when QM/QED were actually written down… 

 

Seems to be more rooted in psychology than photon energy spectra. 

[BQ Octantis]

As we mentioned m earlier discussions, some kinds of variations across individuals have no effect on color matching. The question is not "Is his experience of red the same as mine?" (a very difficult question to answer) but rather "Does his experience of of the image match his own experience of the object?"

I know, right! The first time someone told the the Milky way was green, I laughed at him.

 

"No it isn't! Look!"

 

Full Sky Mosaic in Planckian Color (click for full size)

 

BQ

Edited by BQ Octantis, 09 March 2023 - 09:52 PM.

[BQ Octantis]

When in doubt the photo-spectrometer doesn’t lie… Not difficult to precisely measure the “color” as a function of energy content… If colors weren’t understood to ultra high precision almost none of our technology would function. Suprised there is such debate on something that’s so exquisitely defined and measured without ambiguity in a ways that are repeatable and predictable. Not sure there was as much content when QM/QED were actually written down… 

 

Seems to be more rooted in psychology than photon energy spectra. 

 

Totally true.

 

But absolute color is a vector in a three-dimensional color space that requires all three dimensions (xyY) to specify. However, for this thread we have purposely abandoned an absolute measurement of the relative luminance Y.

 

Now that my Mak 180 is fully calibrated, I have a conclusive answer to the color-matched xy chromaticities in Trapezium.

 

Below are the measured D65 xy chromaticities for 27 10×10 square (3.28 arcsec per side) patches arbitrarily placed in the nebula surrounding the Trapezium star quad. The measurements were produced from the raw data with a calibrated image chain with the CCM optimized for the measured chromaticity range:

 

 

The measured relative luminance is arbitrary.

 

Q.E.D.

 

BQ

Edited by BQ Octantis, 09 March 2023 - 10:15 PM.

[loujost]

I can now photograph spectra, so I can directly see the transmission of my filters, camera sensors, and lenses. This is revealing important differences between camera sensors.

 

Here is my A7R modified for "full spectrum":

 

Attached Thumbnails

• 

[loujost]

Here it is relative to solar irradiance at the earth's surface (blue graph from Wiki media). I've desaturated it to see the Fraunhofer lines better. Below my spectrum is the spectrum that appears as a reference spectrum on the Baader filter packages. This requires a little bit of non-linear stretching.

 

The Fraunhofer lines are labeled. "C" is H_alpha. I see my modified-A7R spectrum shows much darker lines in the UV part of the spectrum. Not sure why. But with this as a reference, I can quantitatively measure the cutoff wavelengths of any filters, to an accuracy of a couple of nm.

Attached Thumbnails

• 

Edited by loujost, 11 March 2023 - 01:13 PM.

[yuzameh]

Wow, what a longtime thread!  And from the look of the OP post count probably the main one used!

 

I'm totally with the OP I'm afraid.  Also the point that people can do what they want.

 

Sky blue Halpha regions nauseate me though.  In fact I find them inherently silly.

 

As RGB as possible.

 

I'm also recently enjoying the handful of people who are posting single line filter images in good old fashioned greyscale.  Otherwise colour them to their wavelength or call them astro-art.

 

Of course if passbands and ranges aren't optical, someone has to play, as hubble often has.  JWST to me has done a decent choice in NIRCAM mapping, galaxies are very analogous to hubble press releases for example.  However, the attempt to get MIRI and NIRCAM to not overlap colours has led to some aesthetically ghastly mid infrared galaxy images.

 

Sometimes you need colour, not just greyscale, say to bring out HII regions amonst the blue stars in sprial arms.

 

However, can you merge 20 colours into an RGB image? How does that work?  I see most JWST stuff are presented as three colour, although some are two colour.

 

Nebulae images are of course the worst abused.

 

I saw someone mention they used an antlia two filter thing, [OIII] and Halpha, and went to the manufacturer's website, there were some example images there from a South Korean chap who appeared to be being faithful to the filters actual colour range, they looked far better than these palette things.

 

Folk can do what they want, of course (as long as its legal and hopefully not antisocial).

 

There's supposed to be a saying "I may not know much about art, but I know what I like"

 

My paraphrase is "I may know nothing about art, but I know what I can't stand"!

[BQ Octantis]

Well then here's a vomitous hurl of true color at the Lagoon for you. I went back and optimized my daylight CCM for the C5 on the cyan-blue-magenta triangle where the Trapezium chromaticities were. It made for some nebula color tone transitions around the Hourglass that I felt I could stand:

 

A Shallow Look into the Lagoon of Truth (click for full size)

 

Measured chromaticities available upon request.

 

BQ

Edited by BQ Octantis, 11 March 2023 - 04:36 PM.

[sharkmelley]

That's right - the overlap between blue and green is quite different.  Take a look at this comparison:

 

SunlightSpectra.jpg

 

The Canon 200D (the plot erroneously says D200) is fairly typical of recent Canon cameras and the Nikon D5300 is fairly typical of Sony sensors.  Take a look at the big difference in the position of the boundary between blue and green.

The practical significance of this Canon vs Nikon difference in blue/green crossover is illustrated by this comparison of M27 (Dumbbell Nebula) courtesy of Jerry Lodriguss:

 

 

Mark

[BQ Octantis]

That's odd…

 

Here's what I got for the Dumbbell from my 600D:

 

 

But that's from back before I knew how to color match.

 

If I redid it today, the stars would be in true Planckian color.

Edited by BQ Octantis, 11 March 2023 - 07:39 PM.

[loujost]

With my cheap spectrometer it is relatively simple to check the spectral response of a camera's sensor. Here are the results of my own cameras, with Fraunhofer lines aligned. (The label "A7R new" just means my newer copy rather than my older copy. It is the original old A7R, not the newer A7R2, 3, or 4.) The two Panasonic sensors did not reach very deeply into the violet, while some other sensors just barely reached H_alpha. The green/blue boundary also varies slightly between them.

Attached Thumbnails

• 

Edited by loujost, 11 March 2023 - 08:41 PM.

[BQ Octantis]

What is your image chain? How are you calibrating it?

 

I have a stock 600D with 8 lenses. The CCM is quite different and unique for every lens. The stock CCM is reasonably close for prints, but not good for precise chromaticity matching.

Edited by BQ Octantis, 11 March 2023 - 08:56 PM.

[loujost]

It is not very precise, I realize. I did use the same Nikon older 28mm-70mm lens for all cameras; I have modified my Pentax K-1 to a Nikon mount to take advantage of that huge lens ecosystem. 

 

I just used the jpgs for this. I should go back and use the RAW images, using the same parameters for each. But the beauty of the spectroscope is that it changes frequency data to spatial data, with the Fraunhofer lines providing a color calibration that is independent of the jpg engine. So my comparisons of spectral ranges across cameras is valid regardless of the color model used. However, the green/blue transition location will certainly depend on the jpg engine.

Edited by loujost, 11 March 2023 - 09:22 PM.

[loujost]

I should add that my monochrome QHY 163 has about the same spectral range as the full-spectrum-modified Sony A7R.

[BQ Octantis]

The in-camera jpegs just use a generic CCM.

 

I can't help but wonder what a custom CCM on the raw would do to the quality of the matching. The matching is pretty spot on across all my lenses (for the Macbeth card, anyway). And given the metameric illuminant failure from all the bloody LED lights these days, I now even have my human subjects hold a card for a test shot!

 

P.S. I'm loving that a7R range!

Edited by BQ Octantis, 12 March 2023 - 07:27 AM.

[BQ Octantis]

Check out the color tones the optimized CCM gave me on the C5 data on the Hourglass and surrounds:

 

 

Those are way better than what I got from asinh, GHT on the camera primaries—or even two years ago with my hand at Roger Clark's "modern workflow". So the CCM version is what I went with for this vomitous art rendition:

 

 

 

BQ

Edited by BQ Octantis, 12 March 2023 - 07:39 AM.

[CDavis1974]

If you walk toward a white wall - does it get brighter?  If you take a rocket to Jupiter - will it get brighter as you approach?  No - it will just appear larger in the field of view.

Doesn't the inverse square law state that it does in fact get brighter?

 

People are talking about etendue, which I will admit that I had never learned about before, but we're not talking about making something brighter than it is, we're talking about making it brighter than it appears to the naked eye.  Obviously we all here have looked through telescopes of various apertures and have been able to see details we wouldn't have otherwise seen at the same magnification at a lower aperture.  Significantly bright objects can show traces of color with a large enough aperture, and again, I think we can all probably say that we've seen this with M42.  I'd certainly never call it "colorful" but I can detect some color.  So how bright exactly?

 

So if M42 is roughly 1,500 light years away, that means that by the inverse square law, it must appear 2.25 million times fainter than it would if we were one light year away.  If we were 15 light years away it would be 10,000 brighter.  We're certainly never going to be talking about popular science fiction depictions levels of brightness, but these bright objects surely must be bright enough to see a faint color in a dark sky at that distance.

 

And someone invoked etendue as an explanation that you can't make the sun appear brighter, but every child who has ever burned ants has seen a small disc of light that is brighter than a similarly sized disc of the sun as viewed from that same position.  The disc isn't "brighter and hotter than the sun" but it is some fraction of the intensity of all the sunlight that falls through its aperture.  Look through a 12" plus Newt and tell me the full moon isn't brighter through that Newt than looking at it across the sky without the telescope.  It's not just spreading the image of the moon out, it is collecting more photons than your pupil normal could.

 

So I think for many objects you could make them have some color given an impossibly sized telescope.

 

At any rate, many of you here are much more knowledgeable about this than I am, these are just some, or at least seem to me, misconceptions that I've seen just accepted.  I'm not really sold on the "true color" notion myself, but it has lead to an interesting discussion nonetheless.

[loujost]

 

P.S. I'm loving that a7R range!

Yes, that is an exciting modification! To make full use of it, though, it requires lenses that focus IR and visible in the same plane, or pure reflectors. There are a few hard-to-find lenses that do this. The early Canon 300mm f5.6 fluorite lens has this property. So does the earliest Nikon 300mm f/4.5 ED (non-IF) lens and the Zeiss/Hasselblad 250mm Superachromat. And a few Takahashi fluorite telescopes. Every now and then. other special limited-production lenses with this property pop up on eBay.

 

Astrophotographers are throwing away a lot of light by using standard camera sensors.

Edited by loujost, 12 March 2023 - 08:26 AM.

[BQ Octantis]

Yes, that is an exciting modification! To make full use of it, though, it requires lenses that focus IR and visible in the same plane, or pure reflectors. There are a few hard-to-find lenses that do this. The early Canon 300mm f5.6 fluorite lens has this property. So does the earliest Nikon 300mm f/4.5 ED (non-IF) lens and the Zeiss/Hasselblad 250mm Superachromat. And a few Takahashi fluorite telescopes.

First world problems…get a quintuplet!

[loujost]

First world problems…get a quintuplet!

One of the myths on CN involve this subject. "IR light needs to be filtered out because it causes star bloat" etc.

 

In fact therer are good reasons for adding IR to our captures.  "Seeing" in IR is vastly better than in visible light. But of course now we are not talking about "true color"

Edited by loujost, 12 March 2023 - 08:51 AM.

[BQ Octantis]

Also, "seeing" in IR is vastly better than in visible light.

Seeing as they launched an entire bloody IR satellite and put it into solar orbit, IR imaging is clearly superior in more than one sense. But photon-phonon confusion means you need cryogenic temperatures for good SNR.

 

But that has nothing to do with color matching or true color—JWST is color blind.

[loujost]

Seeing as they launched an entire bloody IR satellite and put it into solar orbit, IR imaging is clearly superior in more than one sense. But photon-phonon confusion means you need cryogenic temperatures for good SNR.

 

But that has nothing to do with color matching or true color—JWST is color blind.

I was editing my post when you put this up....yes, nothing to do with true color....

 

Thermal IR is well beyond the range that my modded A7R can reach.

[freestar8n]

Doesn't the inverse square law state that it does in fact get brighter?

 

People are talking about etendue, which I will admit that I had never learned about before, but we're not talking about making something brighter than it is, we're talking about making it brighter than it appears to the naked eye.  Obviously we all here have looked through telescopes of various apertures and have been able to see details we wouldn't have otherwise seen at the same magnification at a lower aperture.  Significantly bright objects can show traces of color with a large enough aperture, and again, I think we can all probably say that we've seen this with M42.  I'd certainly never call it "colorful" but I can detect some color.  So how bright exactly?

 

So if M42 is roughly 1,500 light years away, that means that by the inverse square law, it must appear 2.25 million times fainter than it would if we were one light year away.  If we were 15 light years away it would be 10,000 brighter.  We're certainly never going to be talking about popular science fiction depictions levels of brightness, but these bright objects surely must be bright enough to see a faint color in a dark sky at that distance.

 

And someone invoked etendue as an explanation that you can't make the sun appear brighter, but every child who has ever burned ants has seen a small disc of light that is brighter than a similarly sized disc of the sun as viewed from that same position.  The disc isn't "brighter and hotter than the sun" but it is some fraction of the intensity of all the sunlight that falls through its aperture.  Look through a 12" plus Newt and tell me the full moon isn't brighter through that Newt than looking at it across the sky without the telescope.  It's not just spreading the image of the moon out, it is collecting more photons than your pupil normal could.

 

So I think for many objects you could make them have some color given an impossibly sized telescope.

 

At any rate, many of you here are much more knowledgeable about this than I am, these are just some, or at least seem to me, misconceptions that I've seen just accepted.  I'm not really sold on the "true color" notion myself, but it has lead to an interesting discussion nonetheless.

The inverse square law only applies to point sources.  A glowing surface does not change brightness as you move away from it.  It just gets smaller in angle.  The total flux received from a finite object will go down as you move away, but the apparent brightness of a surface does not.

 

Do you get blinded as you get closer and closer to a white wall?

 

M42 is just as bright from here as it is one light year away.  It just takes up a larger amount of the sky - in angle.

 

Similarly, it is just as bright in a big telescope as it is to the naked eye.  In fact it is fainter because all telescopes have inherent transmission losses.

 

By "bright" I mean in terms of the flux per square degree gathered - which is appropriate for talking about the brightness of a white wall, or a patch of nebulosity.

 

The perception of color is complex and involves the eye-brain and can't just be described in terms of photon counts.  I don't know any controlled studies to confirm that people who claim to see color in a faint nebula are truly seeing it.  I would want to see some kind of blind, controlled test where faint images were shown with and without different colors and see if people guess the color correctly.

 

Frank

[loujost]

Similarly, it is just as bright in a big telescope as it is to the naked eye.  In fact it is fainter because all telescopes have inherent transmission losses.

 

By "bright" I mean in terms of the flux per square degree gathered - which is appropriate for talking about the brightness of a white wall, or a patch of nebulosity.

 

 

Frank

I am confused by the telescope case.. A bigger telescope (in terms of aperture) captures much more light from the object than a smaller one, and good camera exposures through that telescope are much shorter (even if we are imaging a big white wall). That's why owls and other night animals have big eyes. Maybe when you said "bigger", you just meant "bigger focal length"?

 

[freestar8n]

I am confused by the telescope case.. A bigger telescope (in terms of aperture) captures much more light from the object than a smaller one, and good camera exposures through that telescope are much shorter (even if we are imaging a big white wall). That's why owls and other night animals have big eyes. Maybe when you said "bigger", you just meant "bigger focal length"?

 

No - this is fundamental and anyone in astro should learn this early.  Telescopes don't make objects brighter - they just make them bigger.  I'm talking about extended objects - not stars.  Stars do indeed get brighter.

 

Any telescope will make objects bigger - and fainter.  Never brighter.  It's very, very fundamental - despite being extremely counter-intuitive.

 

I'm talking about a given human being using a telescope visually.  They are stuck with the eyes they have and the telescope feeds the image into those eyes.  No matter what telescope is put in front - the image will be fainter than if the 'scope is removed.

 

Yes - owls see fainter because they have bigger eyes.  If we could change our eyes we could see fainter also.  But we're stuck with what we have - and putting something in front of them won't help - except in terms of magnification.

 

It's perfectly normal to find this unbelievable and clearly wrong.  But it is quite true - and fundamental.

 

Frank