Different types of filters

Hello, this is my third post, which relates to a post I did a few days ago about filters.

 

I was wondering if any knowledgeable photographers could help me understand the different types of filters that can be used on a camera, their purposes, and how they affect the pictures that are taken in Black and White. Thank you for your time.

Coloured contrast filters will allow same colour light to get through, but block the colours. When you want to change the relative appearance of colours, say red and blue, you can use a red filter to make blue parts appear dark en red very light, or a blue filter to do the opposite.

 

How Black & White Filters Work

 

Apart from contrast filters, there are UV filters that block some more of the UV light film is rather sensitive to (though not much gets through the glass of the lens, so only needed when there is a lot of UV light). And polarizes that selectively block polarized light (mostly light reflected off surfaces).

An optical filter passes its color and and blocks its opposite. A yellow filter passes yellow light (red + green) and block blue light. We mount a yellow filter when using black & white film. Fluffy white clouds are embolden as the yellow filter darkens the blue sky. A red filter does this only stronger. A strong red + under-exposure give the illusion of night. A green filter lightens foliage. A UV filter cuts through distance haze. The UV is commonly used, not for its UV blocking, but to protect our precious lenses. A polarizing filter mitigates reflections from some surfaces.

Not quite. A yellow filter passes yellow light, and blocks also red and green light.

There are no opposite colours. Just colours that are not within the bandwidth of a filter's transmission. A strong red filter only passes rather long wavelengths and blocks shorther ones. A strong bue filter does the opposite, in that it passes rather short wavelengths and blocks longer wavelengths.

Each filter transmits its own part of the spectrum, blocking every wavelight outside that band.

Strong filters (narrow pass) are more (very) effective in blocking light outside its band. Weak(er) filters allow more light outside its band through.

Some filters combine two separate bands of the spectrum. A magenta colour correction filter, for instance, allows both the red and blue parts to pass, blocking the part of the spectrum that lies between these two (mostly green).

 

What filters do is offer a selection of what colour light to use to expose film or sensor, changing the colour balance or the contrast in tone colours are rendered in.

Edited October 30, 2020 by q.g._de_bakker

Not quite. A yellow filter passes yellow light, and blocks also red and green light.

How is that possible?

Do have a look at the transmission spectra of a few yellow filters q.g.

 

Lee filters website is a good place to look.

 

Yellow filters either have a narrow passband between, what might broadly be called, the red and green areas of the spectrum, or they encompass both red and green in almost equal measure and fully block only blue. That's what makes 'em look yellow!

 

Technically, it's clearer to think of a filter in terms of those parts of the spectrum it blocks (absorbs), rather than what it allows through. A water-filter, for example, is meant to absorb contaminants from water, not absorb the water. It should more correctly be called a 'water-contaminant filter' and that way its purpose becomes clearer.

Here's the transmission spectrum of a typical yellow filter on Lee's website.

 

You can clearly see that both the red and green parts of the spectrum are fully included in its transmission and only the complementary - colloquially 'opposite' - colour, blue, is absorbed.

Edited October 30, 2020 by rodeo_joe|1

That, RJ, is because that is not a narrow pass yellow filter.

Yellow is a colour by itself. It is not a mix of red and green. You can see yhe yellow band in the figure on Lee's site. So the answer to how it is possible for a yellow filter to trasmit yellow but block red and green is, quite simply, by transmitting yellow and blocking red and green.

 

The idea of complementary colours comes from tri-colour theory. It is not something found in the spectrum. There is nothing opposite, or complementing, red, blue or yellow. And, again, yellow is a pure spectral colour all by itself.

Edited October 30, 2020 by q.g._de_bakker

Not quite. A yellow filter passes yellow light, and blocks also red and green light.

There are no opposite colours. Just colours that are not within the bandwidth of a filter's transmission. A strong red filter only passes rather long wavelengths and blocks shorther ones. A strong bue filter does the opposite, in that it passes rather short wavelengths and blocks longer wavelengths.

Each filter transmits its own part of the spectrum, blocking every wavelight outside that band.

Strong filters (narrow pass) are more (very) effective in blocking light outside its band. Weak(er) filters allow more light outside its band through.

Some filters combine two separate bands of the spectrum. A magenta colour correction filter, for instance, allows both the red and blue parts to pass, blocking the part of the spectrum that lies between these two (mostly green).

 

What filters do is offer a selection of what colour light to use to expose film or sensor, changing the colour balance or the contrast in tone colours are rendered in.

 

Yellow filters (as well as orange & red ones) are usually long pass filters transmitting light of longer wavelength than yellow (between ~450nm & ~480nm depending on the filter). The only exceptions I know are yellow colour correction filters like the Wratten #6 which (My Kodak Wratten handbook doesn't show the spectra for this filter but specifically states it's not a longpass type).

For those interested the handbook is available on-line here I find it a great reference, but I have a strong technical bent.

 

Even short pass filters designed to transmit only wavelengths below a set value frequently have overtone transmissions well above that value. This becomes a problem for those wanting to record UV images with converted cameras, all the cheaper UV transmitting visual blocking glasses such as Schott U330, BG3 etc. leak considerable NIR something those converted cameras are very sensitive to. Special coatings have to be added to the filters to block these overtones resulting in filters costing hundreds of pounds.

In the case of U330 the overtone transmission reaches about half the transmission of the main band (45% vs 88% on my copy) and transmits over a wider range of wavelengths.

 

Filters rarely block all the light from their excluded wavelengths and never transmit all the light from the transmitted wavelengths (the best transmission I've managed to measure on our works spectrometer is 98%). Transitions from permitted to excluded tend to be reasonably gradual most of my longpass type take over 40nm to go from 1% transmission to 50% transmission, in the case of the 950nm filters this takes around 150nm!

 

Alan's less technical summary is pretty accurate for the actual use of filters, though I feel his description of polarisers is missing out on half their uses. Polarisers set correctly can be used to BOOST reflections something I use every bit as much as their reflection killing properties. :)

 

In everyday B&W photography yellow, orange & red are used to progressively darken skies (making clouds more visible even where a polariser can't manage this)

Green filters are used for skin tones, or for brightening foliage compared to the remainder.

If there are any strong colours present in the scene they can be darkened by using a filter of the complimentary colour.

The best way to explore the behaviour of coloured filters is to convert colour images to B&W using software like Nik's SilverFX, where each colour of filter can be virtually applied at varying strengths simply by playing with sliders.

Well, indeed. But the notion that a yellow filter because it transmits "yellow light (red + green)" had to be corrected. Yellow light is not "red + green". It is yellow.

 

Calling a filter by the colour it is, rather than the colour it blocks, makes life simple, RJ. I rather say red filter than violet, blue, green, yellow, orange filter.

Well, indeed. But the notion that a yellow filter because it transmits "yellow light (red + green)" had to be corrected. Yellow light is not "red + green". It is yellow.

 

You need a lesson in the 3 primary colors of light. There are only 3 colors: red, blue and green light. Other colors are made by mixing only red, blue and green light:

 

Primary Colors of Light and Pigment | learn.

You need a lesson in the 3 primary colors of light. There are only 3 colors: red, blue and green light. Other colors are made by mixing only red, blue and green light:

 

Primary Colors of Light and Pigment | learn.

Yeah... right.

Does the sun know?

You need a lesson in the 3 primary colors of light. There are only 3 colors: red, blue and green light. Other colors are made by mixing only red, blue and green light:

 

Primary Colors of Light and Pigment | learn.

Thats far more down to how WE see light than the nature of light.

Humans have 3 colour light receptors in our eyes, Other animals can have more or less than us I believe there are mantis shrimps that can see 13 colours.

 

There are alos metamers different combinations of colours that LOOK the same to humans even tough they are quite different in composition.

 

At work we have instruments that will differntiate light to much closer than 0.1nm giving rise to thousands/millions of distinct colours of light. I remember some atomic emmision lines having wavelengths specified to less than 1/1000 of a nm (Visual light being roughly 400-700nm).

 

If working with paint or other absorbtive media the human three primary colours are not RGB at all. Red, blue & yellow where those taught at art classes when I was at school, Yellow, cyan & magenta are thos ecommonly used in printers (together with black & often lighter shades of each of the printing primaries)

At work we have instruments that will differntiate light to much closer than 0.1nm giving rise to thousands/millions of distinct colours of light. I remember some atomic emmision lines having wavelengths specified to less than 1/1000 of a nm

In case my memory might be wrong i checked emission lines are typically quoted to 0.0001nm Theres a list of the strongest emission wavelengths for lead here. (Listed in Angstroms which are 1/10nm) :rolleyes:

Edited October 30, 2020 by petrochemist

Thats far more down to how WE see light than the nature of light.

Humans have 3 colour light receptors in our eyes, Other animals can have more or less than us I believe there are mantis shrimps that can see 13 colours.

 

There are alos metamers different combinations of colours that LOOK the same to humans even tough they are quite different in composition.

 

At work we have instruments that will differntiate light to much closer than 0.1nm giving rise to thousands/millions of distinct colours of light. I remember some atomic emmision lines having wavelengths specified to less than 1/1000 of a nm (Visual light being roughly 400-700nm).

 

If working with paint or other absorbtive media the human three primary colours are not RGB at all. Red, blue & yellow where those taught at art classes when I was at school, Yellow, cyan & magenta are thos ecommonly used in printers (together with black & often lighter shades of each of the printing primaries)

But none of that matters a jot to someone that just wants to stick a wide passband Y2(K)* yellow filter on their lens and know that it'll darken a blue sky in a B&W shot.

 

*A yellow filter that most certainly will pass the green and red parts of the spectrum as well as yellow.

The idea of complementary colours comes from tri-colour theory. It is not something found in the spectrum.

It comes from the artists' concept of a colour wheel, where the virtual colours of violet, magenta and purple 'wrap around' to join the red and blue ends of the real spectrum. The CIE horseshoe also embodies this concept, where green is opposite magenta, yellow opposite blue, and cyan opposite red.

 

All conventions are simply constructs of the human mind.

Edited October 31, 2020 by rodeo_joe|1

Yes. Light, however, is not a human convention. Filters have to, and do, deal with that physical reality.

We do too. There is, for instance, no ultra violet on either the artist's colour wheel, the Committee's chart or the tri-colour vision thing. Film however responds to uv light and uv filtering. Same for sensors and ir. If an astrophotographer wants to use hydrogen alpha, the idea of complementary colours is of no use. Sodium vapour lamp light appears a dark yellow, and can be filtered away with either red or blue filters. Et cetera.

Yes. Light, however, is not a human convention.

Very debatable.

The electromagnetic spectrum doesn't 'know' that we humans have visual organs sensitive only to its output between 400 & 700nm in wavelength. It just pumps out energy at all possible frequencies.

 

We choose to call that visible section of the electromagnetic spectrum 'light', and we choose to split that tiny part of the spectrum into what we call 'colours'. We then choose to label those colours with different names, and choose to represent that concept of colour in various types of graphical form.

 

If that's not a construct of the mind, then what is?

 

Another convention is that 'light' only refers to what's visible. Ultraviolet and Infrared aren't strictly light, since we can't see them, and a UV blocking filter appears transparent to us, just for an example.

Edited October 31, 2020 by rodeo_joe|1

If that's not a construct of the mind, then what is?

What we choose to call things are human conventions. That’s different from the thing we call light, which is not a human convention, unless you’re an idealist who believes there’s nothing real outside our minds.

Very debatable.

The electromagnetic spectrum doesn't 'know' that we humans have [etc.].

If only you would see the contradiction in these two sentences.

But we can turn this into a epistemological discussion, or call that something along the line of what some RJ person said about learning basic skills: useless nonsense.

In this context, the latter would be the better.

 

People are mixing up a tricolour vision theory with what light is and what to do to filter part of it (in or out i leave to you).

They, including you, shouldn't.

Yes. Light, however, is not a human convention. Filters have to, and do, deal with that physical reality.

We do too. There is, for instance, no ultra violet on either the artist's colour wheel, the Committee's chart or the tri-colour vision thing. Film however responds to uv light and uv filtering. Same for sensors and ir. If an astrophotographer wants to use hydrogen alpha, the idea of complementary colours is of no use. Sodium vapour lamp light appears a dark yellow, and can be filtered away with either red or blue filters. Et cetera.

Low pressure sodium lights are one of those whose light can be removed fairly effectively using a filter that is reasonable clear to the eye. Neodidyium (red enhancer) filters remove a number of fairly narrow bands that happen to include the prominant sodium D lines.

It's the only example I can think of, that uses reasonably afforable filters to block emission lines :)

What we choose to call things are human conventions. That’s different from the thing we call light.....

I thought I'd clearly pointed that out.

 

Q.g's highly confused argument is that there are no such things as complementary or 'opposite' colours. Except there are. They exist in the models of colour that have been created by both groups of artists and by colour-science consortiums.

 

That a small section of those complementary colours (magenta-purples) don't separately exist in the electromagnetic spectrum is completely beside the point. Those colours are utterly real to our senses, otherwise we wouldn't have invented words for those colours.

 

They are 'imaginary' constructs, but that doesn't make them any less valuable as a conceptual tool.

or call that something along the line of what some RJ person said about learning basic skills: useless nonsense.

Q.g. so far I have resisted lowering my responses to your level of personal insult.

 

However, it's you that's spouting utter nonsense. Your last demented rant directed toward me made absolutely no coherent point whatsoever. It was just a ramble of idiotic bile.

Edited October 31, 2020 by rodeo_joe|1

Let me ask you all one easy question.

 

Is it possible to have a magenta-coloured (minus green) filter?

 

Obviously it is.

 

Case closed!

Misrepresentation, RJ.

My argument is that fiters are dealing with light, in which yellow is a thing in jtself, and not the result of combining two other colours.

Simple as that. Too difficult? Apparently, seeing that you still keep being caught up in colour models. Light, RJ, doesn't care about your models. It is and does what it is and does. Try to understand that.

Let me ask you all one easy question.

 

Is it possible to have a magenta-coloured (minus green) filter?

 

Obviously it is.

 

Case closed!

Only your confused case. Does that make blue or red complementary to yellow?

I just want to think to those who have replied. This has helped me very much.

Is this part of an assignment set by your teacher at school or college ?

 

WW