Exposure correction for B&W filters?

[Drew Bedo]

What is the best way to deal with exposure correction for B&W filters?

[Bob Salomon]

Are you trying to maintain an f stop or an exposure time? Whichever one you want to maintain dictates how to compensate for the filter factor. There is no best way. Aperture is easiest if the factor requires partial stops. But DOF may be important.

[Brian Ellis]

I don't know that there is a universally accepted "best" way. FWIW, my way when using a medium format or smaller SLR camera with a built-in averaging meter of some kind (i.e. not a spot meter) is to meter with the filter on the lens. When using an external spot meter I use filter factors rather than metering through the filter because in the case of filters that alter tones (i.e. yellow, orange, etc.) the color of what's being metered with the spot meter may bear no relation to anything else in the scene. I guess you could meter through a polarizer with an external spot meter since the degree of polarization has no effect on the exposure. I just haven't done it, I use the factor with polarizers too.

These two methods with these two types of meters and b&w filters have worked well for me though I wouldn't claim they're necessarily "the best." I never use incident meters.

[Bob Salomon]

But you better run a rest if you rely of filter factors. They are just a starting point. You want to tune them to your set-up and preferences.

[jim kitchen]

Dear Drew,

I use the manufacturer's published filter factor, and I adjust my f-stop and, or my exposure time as required... :)

That said, and if you want to get very technical, you could expose Kodak's grey scale card under very controlled lighting conditions to determine what the actual exposure factor might be for your favourite filter with a calibrated densitometer. Although, I would bet money that the authors of the filter have predetermined that weighted average factor more correctly, barring any production run upsets or minor quality control failures, compared to what I could ever hope to calculate.

Just my two pennies...

jim k

[bobwysiwyg]

I'm curious, why couldn't one simply take an incident reading without a filter and one with the filter over the meter to determine a difference in exposure needed? Is there some issue associated with different colors and wavelengths?

[jim kitchen]

Is there some issue associated with different colors and wavelengths...?

Dear Bob,

Diffraction might be an issue through the filter's glass or acetate surface, since the incident meter must point to the camera lens, and not the light source... :)

jim k

[bobwysiwyg]

Gotcha, thanks.

[David Schaller]

I meter, with my spot meter, through the filter before I put the filter on the lens. It usually works out to be pretty much the same as the filter factor given, but I always do it anyway.
Dave

[Bob Salomon]

But you better run a rest if you rely of filter factors. They are just a starting point. You want to tune them to your set-up and preferences.

Oops! test of course!

[Bob Salomon]

Is there some issue associated with different colors and wavelengths?

With some CDS cells yes, there was an issue with colors. Hysterisis (spelling?) failure was the problem back in the day. Haven't used that word in 40 or 50 years!

[Brian Ellis]

I'm curious, why couldn't one simply take an incident reading without a filter and one with the filter over the meter to determine a difference in exposure needed? Is there some issue associated with different colors and wavelengths?

I've never used an incident meter so perhaps there's something I'm missing. But what would be the purpose of taking two readings, one through the filter and one without the filter? If you're making a reading through the filter why not just use that reading for your exposure and not bother with the first reading and the subsequent calculation?

As I mentioned in my first message, with an averaging type reflective meter such as is found in most 35mm and digital cameras, and some medium format cameras, I do use the exposure indicated by a reading with the filter on the lens. But I don't do that with a spot meter for the reason I mentioned. Again, this isn't to suggest the way I do it is the best way, it's just the way that has worked for me over a pretty long period of time.

[Brian Ellis]

But you better run a rest if you rely of filter factors. They are just a starting point. You want to tune them to your set-up and preferences.

FWIW I've never done that. The effect of a filter on exposure in b&w photography depends on how much light the filter passes and how much it holds back. And that varies quite a bit depending on the colors in a scene. It also varies with the type of film being used and the film's sensitivity to different colors. So I've always assumed that testing isn't practical. But maybe it is, I just haven't seen a need to do it though I'm sure it would be useful if you were frequently photographing in a controlled environment.

I do make "seat of the pants" adjustments occasionally. For example, if I was using a yellow filter to brighten the yellow flowers in a scene, I might vary the filter factor by half a stop or so depending on how much of the frame the flowers occupied and what the other colors around the flowers were. But that's pretty rare.

Of course the saving grace in all of this is that b&w film has a lot of latitude. It's relatively unlikely that being a half stop or even a stop off because of an incorrect filter adjustment is going to result in an unprintable negative.

[Doremus Scudder]

The basic concept of metering through filters to arrive at the proper exposure is valid. Unfortunately, there are a couple complications: 1) the spectral response of the meter probably will not match that of the film (matching these was the idea behind the Zone VI modified meters, which were supposedly matched to the spectral response of Tri-X), and 2) the film may change speed and contrast characteristics when exposed to light of particular saturated colors.

Fortunately, these can be compensated for with some testing. The approach I use is utilizes a reading taken through my spot meter coupled with "fudge factors" that I have compiled for the films/filters I use. For example, I find that Tri-X (both the new and the old) respond to a #25 red filter by losing speed and gaining contrast. Therefore, when I take a reading through the #25, I automatically add 2/3 stop and develop one zone less. The good news is that it is only the most saturated filters that present the most problem this way. Filters like the #8 or even #11 and #15 seem (in my experience) to not need compensation. However, every film/filter combination is different.

One could run tests for one's favorite film and the most radical filters using a target with a normal contrast range and including a gray card (what I did), or just be aware of the problem in the field and arrive at useful factors by keeping careful field notes (I do that too).

At any rate, IMHO, taking a meter reading through the filter is better than applying exposure factors for a couple of reasons. First, it automatically compensates for the color temperature of the light, so no need to guess how much less/more factor to apply at sunset, or when shooting under tungsten light, etc. Second, reading through the filter with a spot meter is the only way to identify possible mergers of tone that would not be apparent otherwise. Now, you can see that that green tree and the blue sky will be rendered the same shade of gray with that #11 filter you are thinking of using... maybe another would be a better choice.

I've written a couple of other posts on this subject here and on APUG if anyone feels like searching for them :-)

Hope this helps,

Doremus Scudder

[ki6mf]

Ditto what David Schaller and Doremus Scudder said. Meter through the filter and don't worry about compensation.
You could test the compensation by shooting an evenly lit wall then metering each filter to see what the compensation is. When I did this, my yellow was 1/2 stop, my red and orange was 1 stop and my green was 2 stops. Your mileage will vary depending on equipment.

[Brian Ellis]

Ditto what David Schaller and Doremus Scudder said. Meter through the filter and don't worry about compensation. . . .

Just curious, since I do that with an averaging meter in a smaller camera but not with a spot meter. Let's say you're using a 1 degree spot meter and you're planning to use a red filter (maybe to darken the sky, maybe to darken some foliage, doesn't matter). And let's say the darkest important shadow area in the scene - i.e. the thing you're going to meter to determine your exposure) is something red. You'd take your reading through the filter even though you know that's going to give a misleading reading (at least I think it is, since the red filter will allow a lot of light to pass through relative to everything else in the scene)? Or is my premise about the effect of reading something red through a red filter incorrect?

[Brian Ellis]

. . . You could test the compensation by shooting an evenly lit wall then metering each filter to see what the compensation is. When I did this, my yellow was 1/2 stop, my red and orange was 1 stop and my green was 2 stops. Your mileage will vary depending on equipment.

The color of the evenly lit wall and the color of the filter don't matter?

I would have thought that if, for example, I metered an evenly lit yellow wall with and without a yellow filter, and then metered an evenly lit blue wall with and without a yellow filter, the difference between the metered and unmetered readings in the two situations would have been considerably different. Most of the light from the yellow wall would pass through the yellow filter, thereby indicating a need for little exposure compensation with a yellow filter, whereas little of the light from the blue wall would have passed through the yellow filter, thereby indicating a need for a lot of exposure compensation with a yellow filter. So which would I use as my compensation with a yellow filter? Or is there something incorrect in my understanding of how filters in b&w photography work?

[Greg Miller]

Just curious, since I do that with an averaging meter in a smaller camera but not with a spot meter. Let's say you're using a 1 degree spot meter and you're planning to use a red filter (maybe to darken the sky, maybe to darken some foliage, doesn't matter). And let's say the darkest important shadow area in the scene - i.e. the thing you're going to meter to determine your exposure) is something red. You'd take your reading through the filter even though you know that's going to give a misleading reading (at least I think it is, since the red filter will allow a lot of light to pass through relative to everything else in the scene)? Or is my premise about the effect of reading something red through a red filter incorrect?

Typically with a spot meter you will meter the lightest tone and darkest tone, determine how many stops you are working with, and then select an exposure that places your dark and light tones where you want them (using the zone system). Metering through a filter that you will use when making your exposure should not change anything at all. You are just measuring the light that will actually reach the film.

The only possible complicating factor would be if you meter the wrong object because to the eye it appears to be the darkest tone, but another object will in fact be the darkest tone due to the filter. For example, say you have a scene where the 2 darkest objects are a blue and yellow object, and the yellow object is slightly darker than the blue object. But when you use a yellow filter, the blue object will now be darker than the yellow object. If you do not recognize this and meter the yellow object through the filter, then your blue object could lose detail in the exposure. But this is probably not a very likely scenario so isn't worth frettingover.

[Atul Mohidekar]

The color of the evenly lit wall and the color of the filter don't matter?

I would have thought that if, for example, I metered an evenly lit yellow wall with and without a yellow filter, and then metered an evenly lit blue wall with and without a yellow filter, the difference between the metered and unmetered readings in the two situations would have been considerably different. Most of the light from the yellow wall would pass through the yellow filter, thereby indicating a need for little exposure compensation with a yellow filter, whereas little of the light from the blue wall would have passed through the yellow filter, thereby indicating a need for a lot of exposure compensation with a yellow filter. So which would I use as my compensation with a yellow filter? Or is there something incorrect in my understanding of how filters in b&w photography work?

A B&W filter changes the tonality of different parts of the scene in different proportions. A red filter would make all red areas lighter. So if the darkest areas of the scene are red, they might not remain the darkest areas due to the effect of the red filter. So if you are going to use the darkest areas to determine the exposure, you would have to locate for darkest areas while looking through the filter and then take the exposure reading of those areas through the filter. You would do the same if the highlights are used (e.g. for a transparency film) to determine the exposure. In short, look through the filter with the naked eye (without the spotmeter) and locate the darkest or the brightest areas and then take the exposure reading using the spotmeter of these areas through the filter to determine the appropriate exposure - the filter may make totally different areas the darkest and the brightest part of the scene. I hope this makes sense.


// Atul

[Rick A]

I ran tests on all my filters when I first purchased them, and labeled the cases with the filter color and the amount of stops compensation needed. Most filters are what the manufacturer claim, however my polarizer is slightly different, as is my green filter(off by 1/2 stop). I recommend the OP run his own tests for this purpose.

[Bob Salomon]

I ran tests on all my filters when I first purchased them, and labeled the cases with the filter color and the amount of stops compensation needed. Most filters are what the manufacturer claim, however my polarizer is slightly different, as is my green filter(off by 1/2 stop). I recommend the OP run his own tests for this purpose.

There are several methods of making a contrast filter (red, green, yellow, etc.). How they are made can have a big effect on how they perform over time. The most common methods are colored glass (also called "dyed in the mass") these are permanant colors that will not change with time or exposure to light; then there are sandwiched filters with a gel sandwiched between two pieces of glass. This type of filter will fade over time if the dyes used are organic rather then inorganic. A third method uses colored glue between two sheets of glass. Again, depending on what type of dye is used for the color they may also fade over time.

So, if your filters are not solid colored glass then you should repeat the tests periodically. And if your filters are the kind that can shift, do more frequent tests rather then be surprised.