Blue filters and fog

[Q.G.]

As a rough estimate, a yellow is centered at about 450 nm, magenta at 550 nm, and cyan at 650 nm. Of course, this varies with product, unwanted absorptions and half band witdth.

But why do you have to ask me? You are an expert too! Aren't you?

You may want to look here: http://www.kodak.com/eknec/documents/bf/0900688a80316bbf/E7014e.pdf

for confirmation though.

PE

O dear.

If you are able to find magenta at 550 nm, you deserve a Nobel prize or two.

But apparently i'm even a better emulsion engineer than you are, since i can read that graph and know what it says.


But joking aside, we have dragged this thread out long enough.

So just for people still wondering about magenta, it would be good if you too would set things straight.
Magenta is not a spectral colour. So no wavelength.
Cyan and yellow are spectral colours. You can determine their wavelengths.

 

[Monophoto]

- - - if you are photographing a setting with thin fog, will the use of a blue filter accentuate the fog and increase its impact from foreground to background?

Yes - this does work. The effect isn't dramatic, but the use of a blue filter will exaggerate fog.

In movie production, it's also a common practice to use a deep blue filter combined with slight underexposure to create the illusion of darkness . This allows the production crew to film scenes that supposedly occur at night even though they made during normal daylight hours.

 

[Photo Engineer]

Look at page 4 of the reference I gave above at the Kodak web site, and see the spectral absorption of C/M and Y dyes with magenta at about 550 nm.

Magenta, being -G has maximum density at 550 nm on average and maximum transmission at wavelengths on each side. I understand what it says and how it acts alone and in combination with the other dyes. So, on a spectrum of density vs wavelength, you can assign magenta a wavelength, as you can with the other two subtractive dyes.

If you go here: http://en.wikipedia.org/wiki/Magenta you see it defined as a primary, particularly using the color wheel concept. It is defined as such elsewhere. Yellow however, has a dual definition: http://en.wikipedia.org/wiki/Yellow in which it can be defined as a primary or a mixed color due to the way humans perceive it. The latter reference cites work by Bob Hunt of Kodak BTW.

I also recommend "50 Years of the 1931 CIE Standard Observer for Colorimetry" by W. D Wright (Die Farbe 29 (1981), # 4/6, Page 2, fig 1 in which the CIE standard curve was designed essentially around a magenta centered at 550 nm!

PE

 

[Ray Rogers]

Quote:
Originally Posted by Ray Rogers
What happens to white light if you remove the green?

It appears coloured.

You are very tricky!

You know very well that removing the green content would reveal the presence of what you maintain is not there.

How many nm is magenta then?

That question has already been answered.
PE already said it is minus green and so did I when I gave you some approx. wavelengths....



It is clear that I cannot see your POV.

and as you admit, dropping the tricolor goggles does not help us photographically...

But you are correct!

There are different ways of looking at and producing color and vision and I don't doubt you are correct in at least one of those ways!

Unless it has some practical value in photography, however, I don't see why we would want to discuss it...

Anyhow, I will hand the floor over to you guys.

Try not tho hurt each other's feelings too much.

Ray

 

[Q.G.]

Look at page 4 of the reference I gave above at the Kodak web site, and see the spectral absorption of C/M and Y dyes with magenta at about 550 nm.

Magenta, being -G has maximum density at 550 nm on average and maximum transmission at wavelengths on each side. I understand what it says and how it acts alone and in combination with the other dyes. So, on a spectrum of density vs wavelength, you can assign magenta a wavelength, as you can with the other two subtractive dyes.

If you go here: http://en.wikipedia.org/wiki/Magenta you see it defined as a primary, particularly using the color wheel concept. It is defined as such elsewhere. Yellow however, has a dual definition: http://en.wikipedia.org/wiki/Yellow in which it can be defined as a primary or a mixed color due to the way humans perceive it. The latter reference cites work by Bob Hunt of Kodak BTW.

I also recommend "50 Years of the 1931 CIE Standard Observer for Colorimetry" by W. D Wright (Die Farbe 29 (1981), # 4/6, Page 2, fig 1 in which the CIE standard curve was designed essentially around a magenta centered at 550 nm!

PE

PE,

I'm really starting to believe that i am a better emulsion engineer than you are.

The dye forming the magenta image absorbs green.
That's what the graph shows.

So, with all due respect, let me suggest that you read all of those nice reference works.

I can't believe that you quote Wiki, yet fail to register what it is telling you: Wiki.

"Magenta is a pinkish purple color evoked by light stronger in blue and red wavelengths than in yellowish-green wavelengths. (complements of magenta have wavelength 500–530 nm, i.e. green.).[2] In light experiments, magenta can be produced by removing the lime-green wavelengths from white light. It is an extra-spectral color, meaning it cannot be generated by a single wavelength of light, being a mixture of red and blue wavelengths."

The first bit i put in bold is what your graph shows.
I really, honestly, can't believe that you don't know that, that you would believe that it shows that magenta light has a wavelength of about 550 nm (which is green).

The second bit is what i have been saying all along.


Do you really believe what you say here? Really?
If so, i truly can't believe it!

 

[Ray Rogers]

PE,

I'm really starting to believe that i am a better emulsion engineer than you are.

Just for the record... Are you an emulsion engineer?

Ray

 

[Q.G.]

You are very tricky!

You know very well that removing the green content would reveal the presence of what you maintain is not there.

Red and blue? Yes. That's what we would be left with.
When did i claim that that is not there?

How many nm is magenta then?

That question has already been answered.
PE already said it is minus green and so did I when I gave you some approx. wavelengths....

He told us what wavelength green is...

Do some independent research (i.e. don't blindly trust PE). Then maybe you too will know about and understand colour.

It is clear that I cannot see your POV.

and as you admit, dropping the tricolor goggles does not help us photographically...

But you are correct!

There are different ways of looking at and producing color and vision and I don't doubt you are correct in at least one of those ways!

Unless it has some practical value in photography, however, I don't see why we would want to discuss it...

Anyhow, I will hand the floor over to you guys.

Try not tho hurt each other's feelings too much.

Oh, but don't think i am trying to.
I am honestly amazed at how this blatantly non-spectral colour has been assigned a wavelenght by someone of PE's standing. Truly!

So if there's a strength of feeling showing in my posts, it's the magnitude of my amazement, a reflection of the enormous respect i have for PE's otherwise impressive knowledge.
Amazing that something as simple as this has him barking up the wrong tree entirely!

 

[Q.G.]

Just for the record... are you an emulsion engineer?

Ray

No.
Yet my knowledge of colour is much better than that of someone who is.
That's the thing...

 

[Ray Rogers]

Red and blue? Yes. That's what we would be left with.

Excuse me...
You can see the Red and you can see the blue too, at the same time?
As unique colors?

I don't think so!

Not if you are human with normal vision anyway....




I think PE might have made a mistake, (550 nm max is probably absorption max... thus minus green, but even if I am right about that, it doesn't mean he doesn't know what he's trying to say.)

 

[Photo Engineer]

Remember that magenta is a subtractive color and therefore a magenta has an absorption peak at about 550 nm. I did say that earlier, and Ray is correct. I don't want to be misquoted on this! Yellow and Cyan have peaks on each side of this shifted by about 100 nm shorter and longer wrt 550 nm respectively. All subtractive dyes are looked at from the absorption POV while all additive dyes are looked at from the transmission POV. See my post #46

So, to go further, magenta, described in the Wikipedia reference I gave is defined as follows: "In color printing, the color called process magenta, or pigment magenta, or printer's magenta is one of the three primary pigment colors which, along with yellow and cyan, constitute the three subtractive primary colors of pigment. (The secondary colors of pigment are blue, green, and red.) As such, the hue magenta, is the complement of green: magenta pigments absorb green light; thus magenta and green are opposite colors."

Therefore magenta is defined as a primary color in this instance as it is in many others unless one takes snips out of context.

As for using 3 color systems for this, as Wright, who I referred to above, roughly paraphrased from his earlier articles and in his course work says - we are stuck with the 3 color system because it works and that is how the eye works. BTW, he was one of my instructors in the two courses that I took on color. And, applied, my courses work and agree with theory.

So....

PE

 

[Q.G.]

Excuse me...
You can see the Red and you can see the blue too, at the same time?
As unique colors?

I don't think so!

Not if you are human with normal vision anyway....

Yes, seeing red and blue together creates the impression of magenta.
We have been over that before. I have said exactly that before.

That doesn't make anything about magenta existing, like yellow and cyan, as a colour you can assign wavelengths to any less incorrect.
In fact the entire confusion that lead to these incorrect statements about magenta by PE were 'inspired' by the confusion of impressions with realities. I have been saying from the get go that he should stop thinking in terms of tri-colour, since it would (as it has) only lead to errors.

I think PE might have made a mistake, (550 nm max is probably absorption max... thus minus green, but even if I am right about that, it doesn't mean he doesn't know what he's trying to say.)

Yes, he's mistaken in his interpretation of the curve in that Kodak pub.
But a 'yes' too, to that what he is trying to say is incorrect. It started with filtering that would turn pan film into ortho film, where the tri-colour confusion first appeared. And it lasted up to the very point where we are at now.

As far as all the stuff he said about magenta (and yellow and cyan) goes you PE (i hate to talk in the third person when the person talked about is present. Sorry for not changing "he' to 'you' earlier inthis reply!) have given quite strong evidence (though unbelievably so) that you (let's say are rather confused about the matter.
I still believe what i did a few posts before, when i said that we should stop making jokes about it, and just spill the truth: that - even though what you said was very wrong - you really do know how things are.

 

[Q.G.]

Remember that magenta is a subtractive color [...]
So....

PE

So?

So you are trying to save face by pretending we were talking about printing and CRT screens.
I'll not let you.
We were not.

None of that tri-colour (this must be the 100th time i wrote this ) stuff is relevant. (And you did say things that were incorrect even if it would have been relevant).

 

[Photo Engineer]

The misquote of my post #46 is the source of this. I said that 550 is the absorption peak of magenta.

I stress this again.

PE

 

[Photo Engineer]

So?

So you are trying to save face by pretending we were talking about printing and CRT screens.
I'll not let you.
We were not.

None of that tri-colour (this must be the 100th time i wrote this ) stuff is relevant. (And you did say things that were incorrect even if it would have been relevant).

Tri color is relevant as it is the basis of all color photography (subtractive) and all video (additive). We cannot escape this fundamental fact!

Printing of color in magazines as an example, is relevant as they are using subtractive methods and dyes or pigments, and using a photographic original which is tricolor. This is all being viewed by a tricolor human eye!

PE

 

[Q.G.]

The misquote of my post #46 is the source of this. I said that 550 is the absorption peak of magenta.

I stress this again.

PE

O.K.

But i'm still not letting you off that easily.

Who, for instance, misquoted you, and where?
What point were you trying to make when you wrote that bit you say someone misquoted? And how correct was that point?
Why did you say 550 nm when i asked what the wavelength of magenta light would be?
Etc.

And the source of all this, PE, still is the assumption that talk about tri-colour thingies would provide an answer to the blue filter and pan and ortho film thing.

 

[Q.G.]

Tri color is relevant as it is the basis of all color photography (subtractive) and all video (additive). We cannot escape this fundamental fact!

But not (!) relevant to what this thread was about.
I keep telling you this...

Printing of color in magazines as an example, is relevant as they are using subtractive methods and dyes or pigments, and using a photographic original which is tricolor. This is all being viewed by a tricolor human eye!

See?

When did this thread turn into a how do we perceive colours printed on a piece of paper thread?


Anyway, i'm off to bed now.
I'm sure i'll have magenta coloured dreams.

 

[Photo Engineer]

In post #46, I said "A magenta can be represented by a dye or pigment with an absorption peak centered at about 550 nm. A yellow can be represented by a dye or pigment with an absorption peak centered at about 420 nm, and a cyan dye or pigment would have an absorption peak at about 690 nm."

In posts above you (QG and Ray) say

"Originally Posted by Ray Rogers

I think PE might have made a mistake, (550 nm max is probably absorption max... thus minus green, but even if I am right about that, it doesn't mean he doesn't know what he's trying to say.)

Originally Posted by Q.G.

Yes, he's mistaken in his interpretation of the curve in that Kodak pub.
But a 'yes' too, to that what he is trying to say is incorrect. It started with filtering that would turn pan film into ortho film, where the tri-colour confusion first appeared. And it lasted up to the very point where we are at now."

And I said that a true blue filter would render an ortho film (or indeed a pan film) to be only blue sensitive, but a cyan filter would render a pan film to be ortho sensitive and that is where we bogged down.

PE

 

[Two23]

It was a Hollywood trick to use a heavy blue filter to suggest moonlight. I am actually a night time photographer (especially in winter) and can tell you that moonlight is actually more yellow than daylight is. It's not at all blue. As for the cowboy movies, using a heavy red filter did darken the blue sky and make it appear more like night.


Kent in SD

 

[Q.G.]

In post #46, I said "A magenta can be represented by a dye or pigment with an absorption peak centered at about 550 nm. A yellow can be represented by a dye or pigment with an absorption peak centered at about 420 nm, and a cyan dye or pigment would have an absorption peak at about 690 nm."

In posts above you (QG and Ray) say [...]

Yes, PE.
But we have been argueing about magenta and all the rest for much longer, i.e. it's not that.
But i'm sure it's all a big misunderstanding, and today, i think i better let it go.
My apologies for not doing that earlier!

 

[Photo Engineer]

This article is interesting: http://www.biotele.com/magenta.html and is probably the source of portions of this thread, but then step back one level in that site and see: http://biotele.com/ which represents some truly excellent science.

In any event, the early workers on the CIE and Munsell systems defined only 5 primary colors (primary additive and primary subtractive) and this led indirectly to the two color systems. Interestingly, both works showed Magenta and developed the color wheel to show Magenta, which they named Pink.

In any event, Photons can pretty much occupy the same place at the same time but have different wavelengths and this leads to the richness of our color vision. It also explains why Magenta exists. Magenta is well above the alychne and within both CIE and Munsell space which therefore defines it as a "real" color. IR + UV is below the CIE chart and probably near the alychne depending on wavelength which defines that mix as a non-color or a color of no brightness. We can feel it but not see it. UV + Red reaches from the alychne to the bottom of the CIE chart (depending on wavelength as to where they touch) and is seen as a red with out of focus fringes, giving another example of a non-real color in terms of human sight.

PE

 

[Q.G.]

PE,

A lot of this thread exists because of a confusion between what colour is and how we perceive it.

When you look outside, see sunlight falling on the world around us, the light involved consists of the full gamut from short to long wavelengths.
When we talk about how filters work, we're talking about how they block parts of that full spectrum.
When we talk about how filters affect what we record on film, we talk about matching the film's sensitivity to part of the full gamut of wavelengths to how we would like the film to behave in response to that gamut of wavelengths by selectively keeping back parts of that gamut.

We have been over the errors (and have seen the confusion) that regarding all that as a tri-colour affair, instead of the full colour thing it is, brings about.
So once again my urgent plea: do stop thinking about this all in terms of tri-colour (additive or subtractive)! It's wrong, and serves no purpose!


We don't need tricky photons to explain mixes of colour. There are many colours that only exist as a mix of colours. And both these mix-colours and the constituent colours are "real".
But not all of them are spectral. Not all of them exist as a 'constituent' colour in the light the sun pours out over us. Cyan and yellow are. Magenta is not.

And no: that simple little fact is not gleaned from some internetsite.
It is colour 1.0.1., as basic in the 'colour world' as knowing that 1 + 1 = 2 in the mathematics world. Something you are supposed to know. Something people at the CIE certainly know (so stop trying to suggest that the CIE holds a different opinion. They really don't.)
I must say i feel rather insulted that you think i need your references to learn how colour works. Particularly by the more trivial ones! But i can take it.


Finally: do you see how yet again, you confuse how colour works (the trivial URL) with how we perceive colour (the neuro-thingy)?
Would you consider the fact that before we can perceive colour, something out there has to exist? And that that something out there works in a different way, a way shared by emulsions, filters, and all of the rest?
That it is of no importance at all to the question whether a filter can enhance mist, that, say, we can create an impression of green by spinning a disk with only black and white segments?

 

[Photo Engineer]

This discussion exists because on an Analog Photography forum, I am trying to address color in terms of how a photographer and yes, a Photo Engineer would address it, not how a physicist or a neurologist would address it. Yes, I have had to do the matrix math and all of the other things necessary to compute color mixes, and yes, I follow what you are trying to say.

It does not fit well here and muddies the water when another approach is needed. That approach is tricolor. There is no other way to "solve" the problems. Even the matrix math is formed that way in most cases.

And, in the first iterations of the CIE and Munsell systems they didn't get it right. There were ammendments to the standard as knowledge evolved.

PE

 

[Q.G.]

This discussion exists because on an Analog Photography forum, I am trying to address color in terms of how a photographer and yes, a Photo Engineer would address it, not how a physicist or a neurologist would address it. Yes, I have had to do the matrix math and all of the other things necessary to compute color mixes, and yes, I follow what you are trying to say.

It does not fit well here and muddies the water when another approach is needed. That approach is tricolor. There is no other way to "solve" the problems. Even the matrix math is formed that way in most cases.

And, in the first iterations of the CIE and Munsell systems they didn't get it right. There were ammendments to the standard as knowledge evolved.

PE

But never did they ever even hint at the possibility that white light, the thing they wanted to investigate, existed because there was a blue, a red, and a green sun.

Think about that.

That approach is not only not needed. It's way out of place, and frankly, just so much browned tissue paper.

A photographer needs to know what a filter does to the light that is making his film do certain things.

Light that does not exist as a mix of red, blue and green only.

And film that does not react to red, blue and green light only.
Something, i'm sure, every photoengineer should and will know.
Which is also why every photoengineer should know about and treat light as (as you put it) a physicist, i.e. the way it is.

(And before you begin: no. You can replace red blue and green with cyan yellow and magenta, and it will not improve one iota.)

What muddies the water really is your insistance on that tri-colour nonsense.
I don't know what problem you are trying to solve using that tri-colour thing, but what i do know is both that we haven't seen that problem here (tri-colour is completely irrelevant, out of place), and that (as mentioned many times before as well, and as has become apparent in this thread) you are creating more problems insisting that it has some relevance where it has not.

 

[Lee L]

There are three sample landscape photographs in Ansel Adams 'The Negative', pages 116-117 taken by John Sexton with no filter, Wratten #47 blue tricolor filter, and a combination of #25 red tricolor and polarizer filter. There is a relatively nearby hill and a more distant mountain range. Compared to the photo with no filter, the #47 filter (with the standard filter factor applied) produces the distant mountain range about one 'zone' lighter and with dramatically lower contrast within that mountain range than with no filter.

Adams makes a distinction between atmospheric haze and "particulate" haze, i.e. haze caused by smoke or dust. He indicates that particulate haze isn't very manageable with visual-light filters, and are cut more readily with infrared film. On the other hand, he shows clearly that blue filtration will increase the effect of atmospheric haze (which I'm reading as water vapor), especially over distance.

Adams doesn't address the issue of fog directly, but I think once the atmospheric haze is dense enough to reach the level of 'fog', where one is actually in a ground level cloud, the blue filter doesn't do much (if anything) to increase the effect. In other words, the denser the water vapor, the less dramatic the effect of a blue filter in increasing the visual density of water vapor, explaining Ian Grant's experience with close range work in fog.

 

[cowanw]

It seems to me that we are not making photographs with spectral light from the sun so much as reflected light from the subject. Seems to me also that magenta, for example, is a perfectly respectable and real colour as reflected light. It could be enhanced or darkened by appropriate filtration.
I am thinking that the colour theory that PE uses fits my understanding of the reflected light that I think is making my images.
Spectral light influences the colour of the reflected light, but I don't think film cares whether the brides white dress is yellow because of candle lighting or because it's Gramma's old dress in sunlight,but yellow from age. The film sees the reflected light either which way.