How to Think About Color Filters for B&W

[bmattock]

Another Way to Look at Color Filters

I think we tend to get confused about the use of color filters for B&W because the descriptions we read don't address some underlying assumptions which are incorrect. So I was trying to think of a way to present an explanation of color filters that makes sense. This explanation assumes properly-exposed and developed film to begin with.

I think the basic problem is that we think of B&W film as if it were just 'desaturated' color film, and it is not. Color prints say that white things are white - black things are black. Blue is blue, and so on. Is this not also true of B&W film, at least for the 'white' and 'black' things?

B&W film does not record black as black and white as white. What? That's right, B&W film does not know what white is, or what black is. It records only shades of gray based on a sliding scale of relative brightness. Everything is on a sliding scale of relative brightness - the brightest is white, the darkest is black. Everything else is a shade of gray.

What I mean by that is this - in any scene, B&W film records the lightest shade present as 'white' and the darkest as 'black'. Everything else gets a shade of gray, based on it's relative brightness compared to everything else in the scene.

Now, knowing this - if we use a color filter to increase the RELATIVE brightness of a particular color, then everything of that color gets bumped up the scale of relative brightness - it becomes 'lighter' in shade. If it becomes the brightest light in a scene we're photographing, then it gets recorded as white.

The film is recording LUMINANCE, not B&W RENDITIONS OF SHADES OF GRAY. Luminance is brightness - not color, not shade, not tone. Brighter is whiter with B&W film.

For example...imagine a flower bed. Green leaves and red flowers, in this simple case. With straight B&W film and no filters, both the shade of green and the shade of red present may have the same relative brightness - the same luminance - so they will be recorded as the same shade of gray. This may not be what we want.

So, how do we change it? Well, if we put a red filter on our lens, then we are passing more red light and blocking green. Red things move up the chain of relative brightness - and are recorded as 'whiter' than green things, which move down and are recorded as 'darker' than they otherwise might be. If the red flowers are now reflecting the most light in the scene, they record as 'white' on our B&W film. All we are doing is changing the RELATIVE POSITION of the red versus the green by applying a filter so MORE red than any other color gets through.

The same thing happens in reverse if we put a green filter on our lens. The leaves, being green, are now brighter and therefore a lighter shade of gray than the red flowers, which become darker.

With that said, it is easy to understand what the most common color filters are good for - manipulating relationships among colors.

Colored filters lighten their like - and darken their opposite (called their compliment). So red filters make blue skies dark and ominous, yellow filters less so. Green filters block red - and so can have a positive effect on caucasion skin tones in out-of-doors portraits where red or pink blotches and uneven reddish skin tones are common. However, it will darken red lipstick to almost black, and so will have a heavy effect on a portrait of a woman wearing it.

Here is a good resource:

http://www.schneideroptics.com/filte...black_&_white/

I think that should do it. However, by way of passing, I'd also like to address a couple of common questions about the use (or need for) color filters:

1) Can you use color filters on digital cameras? Maybe. Some work, some get confused. Some that can be put in 'B&W mode' are probably more likely to work than those that cannot.

2) Can you just use Photoshop later on to duplicate the effect of a color filter? Yes, if you shot in color originally. If you shot in B&W, or if the image has already become desaturated to B&W, then the individual color information does not exist - you cannot duplicate the effect of say, a red filter, when everything is shades of gray. What's red? What's green? The image no longer contains that information.

3) Do you 'need' to use color filters? No, not in most situations. It is merely another creative option. It is interesting because it allows us to make more distance between colors that would otherwise render as similar shades of gray, which can serve to highlight or isolate a subject in a photo and make an otherwise boring B&W photo 'pop'. Continued overuse of the same filter effects can become tiresome, however. A heavy dark foreboding sky rendered with a dark red filter is cool - but after a couple hundred of them, one begins to feel the need for a visit to the beach.

I hope this is helpful.

Best Regards,

Bill Mattocks

[pshinkaw]

I have noticed tha ta lot of conventional wisdom on using filters for panchromatic B&W photos don't seem to hold up as well as they used to. By used to, I'm speaking about conventions that were commonly understood inthe 1950's and 1960's.

For example, I've noticed that I need yellow filters a lot less now than I did in bygone years to bring out clouds. Has the atmosphere changed? or has the color sensitivity of B&W fims changed over the past 40 years?

Anyone have some old negatives that they can look back to to see if this is true? Perhaps it's just a figment of my imagination.

-Paul

[dnk512]

Thank you for the write-up Bill!

[OldNick]

Bill,

When I first started reading about filters, many years ago, they were frequently referred to in the complimentary sense, such a Orange (Minus-Blue). For those with little knowledge of the color spectrum, this seemed to help understand the effect, i. e., use of the orange filter would move the blue area closer to the black. For some reason, this has stuck with me.

Jim N.

[bmattock]

Originally Posted by pshinkaw

I have noticed tha ta lot of conventional wisdom on using filters for panchromatic B&W photos don't seem to hold up as well as they used to. By used to, I'm speaking about conventions that were commonly understood inthe 1950's and 1960's.

For example, I've noticed that I need yellow filters a lot less now than I did in bygone years to bring out clouds. Has the atmosphere changed? or has the color sensitivity of B&W fims changed over the past 40 years?

Anyone have some old negatives that they can look back to to see if this is true? Perhaps it's just a figment of my imagination.

-Paul

Paul, my first foray into film was in the late 70's, so I can't really speak to the films of the 50's and 60's, but I believe that films are more panchromatic than they were. It took me a long to time to figure out that film does not respond uniformly to increases in light - it is not a straight line OR a logrithm, but rather has a distinct 'shoulder' and 'toe'. Based on this - if the shoulder and toe changed, I could understand how the need for filtration would change as well. But I am not an expert here.

Best Regards,

Bill Mattocks

[bmattock]

Originally Posted by OldNick

Bill,

When I first started reading about filters, many years ago, they were frequently referred to in the complimentary sense, such a Orange (Minus-Blue). For those with little knowledge of the color spectrum, this seemed to help understand the effect, i. e., use of the orange filter would move the blue area closer to the black. For some reason, this has stuck with me.

Jim N.

For someone who is color-blind, such as myself, the thought of color filters for B&W work seemed counter-intuitive for a long time. However, there are not that many colors to remember, and I found that eventually I knew which filter to reach for to get a particular effect, based on past results if not on understanding. I find that I use a medium yellow more than any other filter, but a deep red is my second most-used filter. I seldom use orange or any of the combinations such as yellow-orange or orange-green or what have you. And I personally have never found a good reason to use a blue filter - and seldom a green either.

Best Regards,

Bill Mattocks

[peterc]

Originally Posted by bmattock

and seldom a green either.

A green can be good for portraits. With many films it gives a more natural skin tone.

I have also found a significant difference in the way some of the chromogenic films handle filters compared to traditional B&W emulsions.
XP2 is generally pretty good (didn't handle a dark red filter well though) ... but I had a few rolls of Kodak's offering go muddy when a filter was used.

Good piece on filters Bill .... thanks.

Peter

[pshinkaw]

There was a wonderful Calvin and Hobbs Sunday comic strip about when the world was originally in B&W and when it suddenly turned to color. I feel like that sometimes.

I have many photos taken in the 1960's where I could not seem to get those puffy white clouds to come out on film. When I finally was able to afford some filters, it took a red filter to bring them out to the level I wanted. Now, using Neopan, Tri-X or any other modern film, they pop right out to the level that previously required a medium yellow.

Something somewhere changed. Perhaps the colors of the world just changed.

Anyway, conventional wisdom from my old Ansco and Kodak brochures from the 60's suggest that a green filter will lighten green foliage to make it look more white on film. That will make it contrast more with the sky which should darken as the foliage lightens. At least if the worlds is still the same color.

-Paul

[bmattock]

Originally Posted by pshinkaw

There was a wonderful Calvin and Hobbs Sunday comic strip about when the world was originally in B&W and when it suddenly turned to color. I feel like that sometimes.

I have many photos taken in the 1960's where I could not seem to get those puffy white clouds to come out on film. When I finally was able to afford some filters, it took a red filter to bring them out to the level I wanted. Now, using Neopan, Tri-X or any other modern film, they pop right out to the level that previously required a medium yellow.

Something somewhere changed. Perhaps the colors of the world just changed.

Anyway, conventional wisdom from my old Ansco and Kodak brochures from the 60's suggest that a green filter will lighten green foliage to make it look more white on film. That will make it contrast more with the sky which should darken as the foliage lightens. At least if the worlds is still the same color.

-Paul

And after it rains
There’s a rainbow
And all of the colors are black
It’s not that the colors aren’t there
It’s just imagination they lack
Everything’s the same
Back in my little town
Nothing but the dead and dying
Back in my little town

- My Little Town, Simon & Garfunkel

Best Regards,

Bill Mattocks

[bcs89]

As the result of much reading on the subject (a lot of it here), I felt I had a pretty good grasp of the filter / b&w film relationship - but reading your post has really simplified, or "tidied up" the whole concept for me. Thanks Bill!

[jan normandale]

Bill, keep it up and maybe we'll petition Jorge for 'Bill's corner"..

cheers

[FrankS]

Jan, I think keeping Bill in a corner would be a good idea too. Good thinking.

[Nikon Bob]

I guess it is shown here http://www.schneiderkreuznach.com/ti...tografie_e.htm in relation to a colour wheel. Sometimes seeing a diagram can help.

Nikon Bob

[bmattock]

Originally Posted by FrankS

Jan, I think keeping Bill in a corner would be a good idea too. Good thinking.

With a dunce cap on my head and my thumb in my mouth. I know, I know.

Best Regards,

Bill Mattocks

[vrgard]

Originally Posted by bcs89

As the result of much reading on the subject (a lot of it here), I felt I had a pretty good grasp of the filter / b&w film relationship - but reading your post has really simplified, or "tidied up" the whole concept for me. Thanks Bill!

I just wanted to second Scott's comment above. Thanks very much Bill for putting this subject into terms so easily understood.

-Randy

[FrankS]

Seriously, I do think that sometimes RFF members write truly seminal pieces that should be preserved in an accessible place. Bill has written at least 2 such pieces that i can remember in the recent past.

[Rafael]

Thank you Bill. That was an extremely helpful post.

[oftheherd]

Originally Posted by bmattock

[b]

...

B&W film does not record black as black and white as white. What? That's right, B&W film does not know what white is, or what black is. It records only shades of gray based on a sliding scale of relative brightness. Everything is on a sliding scale of relative brightness - the brightest is white, the darkest is black. Everything else is a shade of gray.

What I mean by that is this - in any scene, B&W film records the lightest shade present as 'white' and the darkest as 'black'. Everything else gets a shade of gray, based on it's relative brightness compared to everything else in the scene.


...

Bill Mattocks

True and yet ...

The hardest thing I had to learn was that b/w photographs generally should have both black and white. That isn't always easy, and is why some take to the zone system so wholeheartedly. Too much gray and they are just muddy. Of course, high key photos may not have much if any black, but generally, the most successful b/w photos should have some of both, or something that easily passes for such.

Anyway, I don't mean to take away from a good post on b/w filters. Thanks for the good post Bill. I haven't done that much b/w in years. When I did more, I tended to use them more. But that was years ago. I needed that reminder to jog some old brain cells, and to make me think about doing more b/w and using filters for just the things you mentioned. Thanks again.

[remrf]

All of this appears to pass completely over the fact that we are not shooting b&w positive film. We are shooting negative film. I would think any attempt at explanation of filtration should include this fact. To jump from real world color to final b&w print without the actual intermediate step of the negative which is what we actually produce gives a partial and possibly confusing description to the newcomer to the field.

[FrankS]

Originally Posted by remrf

All of this appears to pass completely over the fact that we are not shooting b&w positive film. We are shooting negative film. I would think any attempt at explanation of filtration should include this fact. To jump from real world color to final b&w print without the actual intermediate step of the negative which is what we actually produce gives a partial and possibly confusing description to the newcomer to the field.

Unless I'm missing something (this is not an impossibility) Bill has described the affects of the colour filters in terms of the final results in print, so your positive/negative concern has been fully addressed.

[remrf]

Originally Posted by FrankS

Unless I'm missing something (this is not an impossibility) Bill has described the affects of the colour filters in terms of the final results in print, so your positive/negative concern has been fully addressed.

Maybe I missed something. He was talking about brightness of relative color and in fact what is happening is the darkening (greater saturation of the silver halides) of certain wave lengths which THEN produce a lightening on the positive.

[VinceC]

The filters affect how the image is recorded on the negative, which is then turned into a positive print (or scanned). Bill's analysis did address the final image.

His brilliant explanation of relative brightness is crucial to understanding black-and-white photography (also crucial to understanding why sunspots appear to be black even though they're bright white).

[VinceC]

>>Maybe I missed something. He was talking about brightness of relative color and in fact what is happening is the darkening (greater saturation of the silver halides) of certain wave lengths which THEN produce a lightening on the positive.<<

A red filter will cause less blue light from the sky to reach the negative. Bill's point is that saturation doesn't come into play. Only density. The filtered sky will be less dense on the negative and so will print darker on the positive. The Red Stop Sign will transmit all of its photons to the negative, which will be denser, and the stop sign will appear brighter than the blue sky, using a red filter. Switch to blue filter, and the blue sky will appear bright white while the stop sign will be black.

[remrf]

Originally Posted by VinceC

>>Maybe I missed something. He was talking about brightness of relative color and in fact what is happening is the darkening (greater saturation of the silver halides) of certain wave lengths which THEN produce a lightening on the positive.<<

A red filter will cause less blue light from the sky to reach the negative. Bill's point is that saturation doesn't come into play. Only density. The filtered sky will be less dense on the negative and so will print darker on the positive. The Red Stop Sign will transmit all of its photons to the negative, which will be denser, and the stop sign will appear brighter than the blue sky, using a red filter. Switch to blue filter, and the blue sky will appear bright white while the stop sign will be black.

I think we are talking about the same thing using differing terms. Photon saturation of the silver halides (or the degree of saturation which then shows on the negative as varying degrees of density)) is the whole ball game. My point was that as I was reading the explanation it occurred to me that the real intermediate step of the negative image was not mentioned and this could cause confusion to the uninitiated.

[bmattock]

Originally Posted by remrf

Maybe I missed something. He was talking about brightness of relative color and in fact what is happening is the darkening (greater saturation of the silver halides) of certain wave lengths which THEN produce a lightening on the positive.

You are absolutely right, of course. But I was actually attempting to gloss over the technical details in the interest of cutting to the chase. My point here was that when one thinks of the effects of color film (either projected slides or looking at a print) one sees that red records as red, blue as blue, and so on. More importantly, white light produces white prints or projected slides - black produces black. These last two lead to a basic misunderstanding of how B&W film records light.

It is (to my mind, at least) so easy to try to 'match up' the color spectrum as applied to color film and apply it to B&W, especially when one has both black and white points as a reference. One thinks to oneself, "Ah ha! White things in my scene will record as white, and black things as black - everything else must be some shade of gray." And this is the fallacy that makes understanding color filters for B&W use difficult. In fact, the brightest thing in the scene will record as white - regardless of its color. The darkest thing in the scene will record as black - again, regardless of actual color.

The color filter's job, then, is to pass the light of the selected color so that it becomes brighter relative to the other colors present - either because we wish to make that color tend more towards the lighter side of gray, or because we wish the complement (opposite) of that color move towards the darker side of gray - again, relative to the other colors present.

If two colors are present (as in my original thesis) of say red and green, if they are both of the same intensity (luminance, brightness, etc), then they will both record as the same shade of gray. If one wishes to make them appear to be different from each other, a red or a green filter will do the job. Which one to pick is a matter of esthetics.

Can my thesis be attacked for technical accuracy? Sure it can - I've ignored a bunch of things that would make a physics major blush - and I've made a mockery of the actual differences between say, luminance, brightness, and intensity - all of which are technically different things and not interchangeable in a technical discussion. But I think I am successful in illustrating the point that B&W film is different on a fundamental level from color film - it doesn't 'map' to colors like color film does - it uses the intensity of the light to determine what is white and what is black. Understanding this leads to an easier grasp of what color filters are actually good for.

At least, that's my hope.

Best Regards,

Bill Mattocks