Filters for B&W

[Kiwi7475]

[QUOTE=AdamD;1536158]

Why assume that the grad filter is 3 or 4 stops?

It was an example just to illustrate the problem.

If my sky is 4 stops brighter than my foreground, I would like to bring it down to just 2 stops. That means I need to take 2 stops out of the sky.

How do you do that?

I'm gathering a yellow filter, but what I don't understand is, how many stops will a yellow filter darken the sky? 1 stop? What about a yellow orange filter? 2 stops?

I've come to this conclusion, b&w film doesn't tend to need GND filters, but instead color filters. That was confusing to me initially. But now I don't quite understand the relationship between the filter and the impact it has on the intensity of the light. I get the color shift, but not the impact.

Sometimes we may overthink things too much.
A yellow/orange filter will introduce 1.5 to 2 stops loss, a red typically 3 stops, but in the end the manufacturer tells us that or we can find it ourselves with a meter.
Now, it is also a tuned color filter and as such it will change the transmission depending on the wavelength. If I have a red rose and use a red filter, I will lose ~3 stops, the rose will appear basically white, and any greens will get darker. With a green filter maybe I lose ~1 stop and the rose will appear dark and the leaves fairly light.
You can’t make a general rule because it depends on what your camera is looking at. Even the same red filter may give you a different effect for the sky at a different time of the day!
So I’d suggest to just account for the nominal filter loss and consider that generally you will darken/lighten according to what you have in front of you based on general rules.

[tuco]

... What I mean by balance is to reduce the difference between foreground to sky to 2 stops from say 3 or 4 stops. In this case, I need to darken the sky by 1 or 2 stops to balance the exposure. Am I making sense? ... If so, how many stops will a yellow filter bring down a blue sky? How about yellow orange? If not, how would you accommodate a 3 to 4 stop variance between sky and foreground? For me in Arizona, that's a normal day; a ton of sunshine!!

Your 'balance' from using colored filters will come in the form of tonality changes. The filter will drop the EV values of the entire scene as already noted. But darkening say the sky a little in tonality with a colored filter may look like you've reduced the exposure for that area for instance.

There a lot of factors besides a colored filter that determine how dark your sky will get when using them. How blue the sky is and where you placed your middle grey exposure will affect the outcome, for example. If you are in the city and use a yellow or orange filter, chances are a the filter will not darken your sky as much as being up high in the mountains in a remote area. I can't give quantitative values except to say a red filter will darken the sky more than a orange and an orange will darken it more than a yellow. Consult a Color Wheel to see why. Colored filters lighten their color and darken the opposite color as seen on a color wheel and to a lesser degree with adjacent colors the further away from the filter color you get.

Graduated filters can help but they are not always the answer. If you have say tall trees or building protruding into the sky, the grad ND will darken them too. With BW film, you can compress your highlights when developing the film. That is how you can tackle your problem. And you can use the colored filters to aid in separating colors that would otherwise reproduce to a similar shade of grey without their use.

[Doremus Scudder]

I beg to differ...

I'm sure I could show you a few where you'd have no idea I used a GND filter. Yes trees shooting up in the sky can be problematic though.

We perceive colored skies differently than b&w film renders and especially with regard to dynamic range, so, I would posit that oftentimes a well-placed GND in certain situations will look more "true-to-life" than a straight shot.

At the end of the day the OP must try things for themselves as likely he has his own sensibilities in this regard.

Why assume that the grad filter is 3 or 4 stops? ... Usually a 1 or 2 (tops) stops is more than sufficient (unless you have extreme conditions such shooting a sunset straight into the sun). They also make the scene more natural to the eye because of how we perceive the color of the sky. And generally one uses “soft” rather than “hard” grads — also helps with making protruding elements not obviously “darkened”.

If your shooting slide and there’s a sky, you’re almost obligated to use one as they only have 5-6 stops of dynamic range.

Finally you can dodge during printing or even easier retouching it after scanning the negative and that removes any artifacts on protruding elements, and balances the image.

Corran and Kiwi,

I guess I shouldn't be so offhand with my remarks It's just that I see so many instances of badly-used GND filters that I have an allergic reaction whenever I see another (and I'm including cinematography, television shows, Peter Lik and his ilk and the horrible background slide show on my Chromecast as well as art photography).

And, the reference to "3-4 stops" was from the OP and in response to how color filters work; I wasn't referencing GND filters.

Any tool in the hands of a sensitive craftsman will deliver good results. And, I'm aware that transparency film poses a particular problem due to its limited dynamic range if one wants the transparency itself to be the finished product. I imagine that if I were still shooting transparency film, I'd likely look into using GND filters... However, with negative materials, black-and-white in particular, and transparencies that get scanned and post-processed, judicious burning, etc. can do the job as well as a GND. And, I really hate overly-burned in skies (with the attendant dark trees, etc.) as much as I do indiscriminate use of GND filters.

And, it should be obvious when I'm posting something that is opinion rather than fact; of course the OP and anyone else is welcome to disagree, even when I'm right

... I'm gathering a yellow filter, but what I don't understand is, how many stops will a yellow filter darken the sky? 1 stop? What about a yellow orange filter? 2 stops?

I've come to this conclusion, b&w film doesn't tend to need GND filters, but instead color filters. That was confusing to me initially. But now I don't quite understand the relationship between the filter and the impact it has on the intensity of the light. I get the color shift, but not the impact.

Adam,

How much a yellow filter will darken a blue sky depends on how blue the sky is. Same goes for red and orange. Read my previous post again and peruse the filter catalog I linked to. All the answers are there. Please, if you want us to help you, make the requisite effort to make that possible.

Best,

Doremus

[Alan Klein]

I think the question Adam has is as follows. If let's say we use a orange filter to darken the sky. The filter manufacturer says add two stops. Well, let's say the entire picture has a lot of blue in it. Would we need to expose even more because the filter is blocking light? O do we keep it at 2 stops. Or let's say the opposite. The scene actually has a lot of orange which will pass the color. Should we reduce the stop from 2 to let's say 1 or 1/2 stops since the orange filter is passing more of the orange color?

In other words, does the colors in the scene affect the amount of stops? Actually, that's a question I have in any case.

[tuco]

That makes sense, Alan. And it seems so variable how would it be quantified. Maybe it has been done.

[Doremus Scudder]

I think the question Adam has is as follows. If let's say we use a orange filter to darken the sky. The filter manufacturer says add two stops. Well, let's say the entire picture has a lot of blue in it. Would we need to expose even more because the filter is blocking light? O do we keep it at 2 stops. Or let's say the opposite. The scene actually has a lot of orange which will pass the color. Should we reduce the stop from 2 to let's say 1 or 1/2 stops since the orange filter is passing more of the orange color?

In other words, do the colors in the scene affect the amount of stops? Actually, that's a question I have in any case.

Alan,

You've uncovered the heart of the problem here. The manufacturers lie about filter factors!

Really, like most things, the closer you look at this topic, the more complex it becomes.

First, filter factors are approximate compensations to get a similar exposure for neutral-colored objects. They change around in reality based on the light source (daylight vs tungsten vs warm light at sunset vs skylit shadows, etc.), the spectral sensitivity of the film (pan, ortho, extended red sensitivity, etc.) and the colors in the scene (use a red filter to photograph a scene with nothing but saturated blue and green and you'll get just about nothing... regardless of how much you expose).

Let's take your example: We use an orange filter and hope to darken the sky using panchromatic black-and-white film. We compensate using the filter factor given by the manufacturer (two stops in this case) and we assume that the lighting (daylight here) is the same as what the manufacturer used to come up with the filter factor.

A grey card in the scene would be rendered the same middle grey as it would if we had made the photo without a filter and at the original exposure. That's what the filter factor is based on.

The blue sky will be darker in the final print (lighter in the negative) than a non-filtered photo. How much darker? That depends on how saturated blue the sky was at the time. Milky-blue sky at the seaside = less darkening. Jet-blue sky at 10,000 feet in clean alpine air = a whole lot of darkening. How do you determine how much darkening you will get before you make the photo? Many just guess, based on experience. I meter through the filter knowing that I'm not going to get a 100% accurate reading and adjust my results with the data I have from testing and taking notes; still not 100% accurate but better than guessing.

You see, coming up with a precise measurement of exactly how a given filter will affect a given object is difficult. We photographers in the field don't have the tools to make such measurements and have to guess, guesstimate, do our best with inadequate measurements, use our experience, etc., etc. But back to our scenario...

Now how about those shadows lit by blue skylight? They are going to be rendered darker in the print too since the light source for them is a whole lot bluer than daylight. How much bluer? That depends on the sky... Shadows lit by deep blue skylight may end up being inky-black (detail-less in the negative) if we aren't aware and careful when making the photo.

Now how about those orange and red leaves on the trees over there? They will be rendered lighter in the final print (darker in the negative) than a non-filtered photo. How much lighter? Depends...

Should you change your filter factor depending on the colors in a scene? Well, likely not, since we choose a filter for its effect. If you don't want blue things to be rendered darker, then don't use an orange filter and vice-versa. Evaluating the scene and choosing the filter that is going to give you the results you desire is an art in itself, and a not-really-quantifiable one

Should you change your filter factor depending on the color temperature of your light source? Absolutely! Film manufacturers used to give filter factors for daylight, 2700K tungsten and 3400K tungsten. Factors were different for different films for each of these sources due to the spectral sensitivity of the film. Nowadays, this seems to have slipped into the background. The laws of physics, however, haven't. Careful workers search out the data and/or do tests. Still, some scenarios are problematic. What's the right filter factor for an orange filter used to photograph something lit by orange light from a sunset? Your guess is as good as mine. Certainly, less exposure compensation will be required since the light is the same color as the filter. But my meter may not give me an accurate reading in that colored light either, so I'll meter and err on the side of overexposure to be sure. It's just not an exact science sometimes.

See what I mean?

Let's look at a couple of other examples: Let's photograph a stop sign with a deep red filter. Result in the final print? A white sign with no legible print on it at all.
The same sign with a deep green filter? Bright white lettering on a dark, almost black, background. Yellow leaves against the blue sky with a yellow filter? = lighter leaves and darker sky. How much? See above.

Best,

Doremus

[JMO]

To the OP, another thing to keep in mind is that if you are in the field and using a spot meter to determine your exposure, you can meter the scene through the actual colored filter, so you can determine what effect the filter will have on your exposure in different parts of the composition. When I do this, I have found that the mfr's recommended filter factors are usually a bit too strong, so if you follow them you'll often be overexposing somewhat. But you're in control, and with experience and all of the good advice above, you'll become more confident.

[Doremus Scudder]

To the OP, another thing to keep in mind is that if you are in the field and using a spot meter to determine your exposure, you can meter the scene through the actual colored filter, so you can determine what effect the filter will have on your exposure in different parts of the composition. When I do this, I have found that the mfr's recommended filter factors are usually a bit too strong, so if you follow them you'll often be overexposing somewhat. But you're in control, and with experience and all of the good advice above, you'll become more confident.

I do this all the time (and recommended it to the OP in my above posts...), but you have to keep in mind that the film's spectral response is not the same as the meter's. So, especially with the stronger filters, you're going to get inaccurate readings. These can be corrected by doing a few tests and finding factors for a given meter/film/filter combination. Note that this is also film-specific; I have different factors for #25 filters with Tri-X and 400TMax for example.

To complicate things more, the contrast index for a given film developed for the same time will change depending on the filter used. Again, stronger filters have the most effect. For me, 320Tri-X gains about a Zone of contrast with a #25 filter and proper compensation. I have to develop N-1 to rein things in. 400T-Max goes the other way...

Best,

Doremus

[Peter De Smidt]

I recommend that the original poster gets a yellow/green filter and an orange filter. Use them and evaluate.

For example: https://hoyafilter.com/product/x0_yellow_green/

and

https://hoyafilter.com/product/ya3_pro_orange/

If that's not enough, you can always get a red filter.

[LabRat]

I think the question Adam has is as follows. If let's say we use a orange filter to darken the sky. The filter manufacturer says add two stops. Well, let's say the entire picture has a lot of blue in it. Would we need to expose even more because the filter is blocking light? O do we keep it at 2 stops. Or let's say the opposite. The scene actually has a lot of orange which will pass the color. Should we reduce the stop from 2 to let's say 1 or 1/2 stops since the orange filter is passing more of the orange color?

In other words, does the colors in the scene affect the amount of stops? Actually, that's a question I have in any case.

This is how filters work on optics.... Same colors pass with very little cut, but complementary colors are where the filter factors are...

The next problem is Pan film has a range of color sensitivity that varies slightly across the visual spectrum, so factors can vary between different films, so this is where testing for specific filter factors comes into play... You do it to account for different films, not the filter's density...

Depending on the colors in a scene, you can reach the point of contrast that can exceed the range of the film, so try to make intelligent, informed choices when selecting a filter if a full range result is desired...

Steve K