Filters on b/w film

[Alan Edward Klein]

Musings about filter factors:

The logical "standard" for a factor for a given filter would be the extra exposure needed to render a mid-tone neutral gray the same density on the film with and without the filter. Other colors would have more or less density on the negative depending on the degree to which they are transmitted (blocked) by the filter. E.g., a red filter, which transmits mostly red light, would need enough extra exposure to transmit enough of the red in the neutral gray to achieve the same density in the negative as there would be without a filter. Other colors would fall where they may.

This factor, of necessity, would vary to differing degrees from film to film, depending on the spectral response of the films, as well from meter to meter, for the same reason. Furthermore, the color content of the light source would make a significant difference in determining factors. (Black-and-white films used to come with tech sheets giving factors for both daylight and tungsten, for example).

So, already, it seems apparent that you can only have an accurate filter factor for a given meter, film and light source. Now take into account the reality of metering in the field. Sure, we can take a gray card with us and use that, but many don't and it would often be impractical. So, we end up metering things that aren't neutral (i.e., don't have an even spectral reflectance). That can easily wreak havoc with filter factors; just try metering that deep blue shadow in an alpine scene to base your exposure on and then add a red filter and the published filter factor. You've just based your exposure on mostly blue light and then blocked all of that out with the red filter and only given an extra three stops to retain some detail in the shadow. Featureless, disappointing shadows are sure to be the result. Or vice-versa; take a reading off that red barn illuminated by the setting sun and then add your red filter. Oops, overexposure! (Which, however, isn't so bad as underexposure).

And, even if we can arrive at a factor that renders spectral-neutral objects at the desired density on the film, determining how much densities from colored objects will be affected is a crap shoot. Usually, we're just guessing: "darker" blue sky (how much darker?). Lighter foliage? How much? (and do we even take into account that although leaves often appear green, that there is a lot of red in there too?)

It would be technically possible to make a meter that had the same spectral response as a given film. Then a simple reading through the filter would suffice for both determining exposure and knowing how colored objects would be rendered. That was the idea behind Zone VI modified meters. I don't think it worked that well though, for whatever reason (I've had two Zone VI modified meters over the years and simply reading through the filters, especially the stronger ones, didn't work as advertised.)

Add to all this the fact that film emulsions react to different colors of light with varying sensitivity and contrast and the plot thickens even more. A deep red filter might end up making one film contrastier and another less contrasty when developed to "normal."

Published filter factors are the manufacturers' best recommendation for getting good results, but they are, of necessity, approximations based on general conditions and (hopefully) a safety factor. One can do all the necessary tests to arrive at better filter factors for the conditions, equipment and materials they use regularly. Many do this (I have). However, the results of such tests are also just guidelines/approximations/educated guesses for compensation, even if more refined.

The best we can do is understand how filters work and how the system reacts to them and be aware of what conditions might require modifying the factor (and how much in which way). We can go down that rabbit hole as far as we like. We can also bracket in questionable situations, not only exposure, but using different filters too. This is especially important when learning in order to gain the experience in using filters that will help us to refine our filter choice and to modify the factors we use.

Best,

Doremus

So should I use the filter manufacturer's suggested factor? That's what I do. The rest is too speculative.

 

[David Lindquist]

It was an attempt at humor, that apparently fell flat.

I thought it was funny.

Not sure Adams used the red filters all that much. He seemed to favor the Wratten #12 based on his books (From My Camera in the National Parks to Examples) wherein he gives technical details on the photographs. He did famously use a Wratten #29 for "Monolith, the Face of Half Dome."

Realizing that for some reason I came to this thread late and that this post to which I just replied, was early.

In my earlier years i thought that if yellow was good, orange was better and red was best (for sky) then I moderated a bit. Last fall I bought a B+W 090, said to be equivalent to a Wratten #25, haven't used it yet.

David

 

[logan2z]

I thought it was funny.

I'm glad at least a couple of people did. Sometimes the forum seems a bit too, um, Sirius

 

[Sirius Glass]

I'm glad at least a couple of people did. Sometimes the forum seems a bit too, um, Sirius

Siriusly?

 

[Doremus Scudder]

So should I use the filter manufacturer's suggested factor? That's what I do. The rest is too speculative.

If you're not into testing to find your own factors, then using the manufacturers' recommendations is a great default compensation. Be aware of the situations that affect the filter factor (e.g., blue-lit shadows, warm light at sunset, etc.) and adjust or bracket those situations.

As you gain experience, you'll find you automatically know what to do. Still, there will always be those times when you may want to make more than one exposure with different filters and/or exposures.

Best,

Doremus

 

[AgX]

So should I use the filter manufacturer's suggested factor? That's what I do. The rest is too speculative.

Using the manufacturer's filter factor is as speculative, as you then assume the manufacturer just had the same idea in mind concerning colour reproduction as you intend when they established that factor.

 

[Raghu Kuvempunagar]

Is there any relationship between reciprocity failure and color of light? I mean, is it possible that reciprocity failure can happen for a specific film earlier or later depending the color of light?

 

[Alan Edward Klein]

Using the manufacturer's filter factor is as speculative, as you then assume the manufacturer just had the same idea in mind concerning colour reproduction as you intend when they established that factor.

Good point. I also forgot that some film data sheets provide filter factors for their film which you ought to follow. Also note that the factors are different for tungston lighting from sunlight.

 

[AgX]

I did not mean film spectral sensitivity nor light spectral distribution, though that plays a part too, but basically what part of the opject you want to remain in rendering and what you want to change. This has to be clear in first place to decide whether to apply a stated filter factor or not.

 

[Pieter12]

Good point. I also forgot that some film data sheets provide filter factors for their film which you ought to follow. Also note that the factors are different for tungston lighting from sunlight.

I am not certain, but unless a filter has a Wratten number, specific colors and tints for black and white are determined by the filter manufacturer and might vary, so a film spec sheet is a starting point.

 

[Sirius Glass]

I am not certain, but unless a filter has a Wratten number, specific colors and tints for black and white are determined by the filter manufacturer and might vary, so a film spec sheet is a starting point.

Exactly. The manufacturer has tested their filters, while the datasheet are using generic filters. Even if a filter factor is off, it will be off by less than half an f/stop and that is well within the range of negative films.

 

[Doremus Scudder]

Is there any relationship between reciprocity failure and color of light? I mean, is it possible that reciprocity failure can happen for a specific film earlier or later depending the color of light?

I've found a relationship between the color of light and contrast. For example, 320Tri-X seems to need about an N-1 development when used with a #25 filter, all other things being equal. I haven't tested extensively to find whether this really is a characteristic of the emulsion or what, but it doesn't happen with TMY, so I would assume so.

It would be interesting to do a few shots at a long exposure time and with, say, a #25, a #47 and a #58 filter - adjusting aperture as needed to get the same exposure time - and see what the results were.

Best,

Doremus

 

[Doremus Scudder]

Using the manufacturer's filter factor is as speculative, as you then assume the manufacturer just had the same idea in mind concerning colour reproduction as you intend when they established that factor.

As I mentioned before, manufacturers' factors are calculated to render a neutral gray card the same density both with and without the filter. How other colors are rendered depends on the filter color, the saturation of colors in the scene, spectral sensitivity of the film, etc. It would be impossible to come up with factors based on those widely-variable parameters.

It is also almost impossible to know how the differently-colored objects in a scene will be rendered on film by simply applying a filter factor, however it is determined. The only real way to do that would be to take meter readings through the filter with a meter that had spectral sensitivity matched to the film you were using. Zone VI tried that with their modified meters, ostensibly matched to Tri-X. I'm not sure that worked all that well.

I read through my filters all the time, and I've developed a set of approximate "fudge factors" for my meter with the films I use. It is far from 100% accurate but the best way I know to at least get an idea of how colored objects will be rendered on film (heck, the same variables apply to unfiltered images too, which we use our meters for all the time).

How exactly do you determine what densities colored objects will have in the negative/final print when you use filters? In other words, what do you do that is less speculative than applying the filter factor?

Best,

Doremus

 

[Doremus Scudder]

Don't sweat it too much. There are too many variables + unknowns to be precise about it. ...

That about sums it up

 

[AgX]

As I mentioned before, manufacturers' factors are calculated to render a neutral gray card the same density both with and without the filter.

I already hinted at this before. But as I also hinted at, to apply the filter factor only makes sense if one wants to render a neutral grey the same after applying the filter. But this is not necessarily the case, as rendering a certain colour the same as before may be a prerequisite in filter application.
Not understanding this I see as the main fault in considering filter application.

 

[Pieter12]

I generally use filters in black and white for specific purposes--usually to enhance skies and clouds or foliage. That pretty much has me using an orange or green filter from the same manufacturers for my most-used lenses. I have tested both with my preferred film--HP5+--and know the filter factor for those combinations. It's not a difficult exercise and can give you a good starting point for continuing the process if necessary when making darkroom prints.

 

[KenS]

I got it. Was grocery shopping so could read the post while the deli person fed me baloney and cut the cheese. When I use a dark red filter I automatically have a beard and a big view camera!

But... I'll bet you 'forgot' the 'three legged support' onto which you use to keep the camera from shaking (or falling due to 'gravity') for that 8 second exposure

Ken

 

[Sirius Glass]

Don't sweat it too much. There are too many variables + unknowns to be precise about it.

Use the suggested factor from the filter or film manufacturer (depending on the type of filter) and some judgment which comes from experience. When in doubt give a little extra exposure and bracket.

A basic understanding of light/colour is of course helpful.

 

[Doremus Scudder]

I already hinted at this before. But as I also hinted at, to apply the filter factor only makes sense if one wants to render a neutral grey the same after applying the filter. But this is not necessarily the case, as rendering a certain colour the same as before may be a prerequisite in filter application.
Not understanding this I see as the main fault in considering filter application.

And just how do you go about rendering, say, bright red at density X on the negative when using a red filter? Or green or blue or whatever?

I can't see an easy or repeatable way to do that, especially given that objects that appear the same color to the eye often have vastly different spectral characteristics.

What's your method?

Doremus

 

[AgX]

Of course that only goes by approximation. Let's say, a red object will be rendered "about" the same with a red filter as without. If one aims at such one would leave aside the filter factor given for that filter.

 

[Doremus Scudder]

Of course that only goes by approximation. Let's say, a red object will be rendered "about" the same with a red filter as without. If one aims at such one would leave aside the filter factor given for that filter.

So you're just guesstimating like the rest of us! Based on a knowledge of color theory and using experience, of course, but still...

Best,

Doremus

 

[Pieter12]

Of course that only goes by approximation. Let's say, a red object will be rendered "about" the same with a red filter as without. If one aims at such one would leave aside the filter factor given for that filter.

I would think something is astray.

 

[BrianShaw]

I would think something is astray.

That is why I left the conversation days ago.

 

[AgX]

You both seem not to understand what I have repeatedly been saying.

As I said already at post #5 , the effect of a filter is dependant on several factors. Put in here the "approximation" from my post above.
However this is something different from blindly applying a given filter factor independent of which object one wants to render the same.

There is common misunderstanding of the meaning of Filter Factor.

 

[BrianShaw]

That is correct… I haven’t a clue what you are talking about. Your thesis doesn’t jibe with my experience. Yes, there are many factors but they do not cause such extremes as you claim. Not using a filter factor on a light yellow… sure; with a red filter… not likely. Using a slightly different factor… sure, depending on those many factors.