Filters Drive Me Crazy

[Don Wallace]

The nice thing about standards is that everyone can have their own. This seems to be the approach of those folks making filters since there are so many different naming systems. I also find filter factors mystifying. I have in my hand right now a b+w orange filter (040 - or perhaps you know it by some other number or letter) and written on the ring is "4x" which I am guessing is the filter factor. This means that I open up by two stops. But if I use my spotmeter on a neutral subject to check this, with the filter right up against the spotmeter, I get only 1 stop difference. I have a Hoya light yellow Y (or K2) filter which supposedly has a filter factor of 2, or 1 stop, yet checking with the meter, the needle moves only 1/3 stop. Finally, I have a Kenko YG filter (which apparently doesn't exist because I can find no reference to it anywhere). It is light green and when I bought it (second hand), someone had written on the box "FF=5." But again, with the meter, I get only 2/3 of a stop difference.

So, here is the question. What do I believe: the filter factor, or the light coming through the filter? Why is there such a huge difference?

 

[keithwms]

A filter factor is just a rough recommendation based on some assumptions about colour temp of the dominant light source and a typical film's spectral sensitivity.

If you can meter TTL then by all means do it. But even that is no guarantee: ultimately, you have to convolve the meter reading through the lens with the spectral sensitivity of the film. For example, if you meter TTL with a red filter and your film has low red sensitivity, then... the filter factor you need will be larger than the meter reading. Another extreme example: many meters can't meter through the opaque IR filters (because a hot mirror is cutting out the IR light), so in that case you have to build on your (and other people's experience).

Note the confusion about filter factors: if the stated number seems to be a multiple of 0.3, then it's probably using the logarithmic scale, in which 0.3=1 stop, 0.6=2, etc. If I see FF=5 then I'd guess it's meant that you should expect to expose at around EV+5 stops. But by all means meter TTL (or TTF!) if you can!

 

[Don Wallace]

Keith, thanks for the fast response. I don't have TTL but I do have a spot meter. My plan for today was to get outside and mess with filters a bit since this is an area of weakness in my knowledge (but I guess you figured that out, lol). Perhaps I should shoot several frames with the same subject and filter, varying the exposure between the filter factor and the reading I get through the spot. Film is still relatively cheap, especially if I do the tests on 120 rather than 4x5.

 

[keithwms]

Yep sometimes I just spot meter through the filter, comparing values with and without the filter for a reality check. Not that the discrepancy can vary depending on whether you're aimed at the sky or a neutral grey object or your foot or....

 

[John Koehrer]

As keith said above, the difference you see is from the sensitivity of the film vs that of the meter.
The filter factors are typically given in factors of .5 so a 2X is one full stop, 4X is two stops 5X would be 2.5 stops. The log factors are seen with the Wratten gels and are a closer tolerance for exposure.
Each color filter may cause your meter to respond differently to the factor. Because of this difference you can't rely on the meter to give accurate exposure. It is a starting point and requires some investigation on your part.
I've found a good guide in Simmons' book "Using the View Camera"

 

[wogster]

The nice thing about standards is that everyone can have their own. This seems to be the approach of those folks making filters since there are so many different naming systems. I also find filter factors mystifying. I have in my hand right now a b+w orange filter (040 - or perhaps you know it by some other number or letter) and written on the ring is "4x" which I am guessing is the filter factor. This means that I open up by two stops. But if I use my spotmeter on a neutral subject to check this, with the filter right up against the spotmeter, I get only 1 stop difference. I have a Hoya light yellow Y (or K2) filter which supposedly has a filter factor of 2, or 1 stop, yet checking with the meter, the needle moves only 1/3 stop. Finally, I have a Kenko YG filter (which apparently doesn't exist because I can find no reference to it anywhere). It is light green and when I bought it (second hand), someone had written on the box "FF=5." But again, with the meter, I get only 2/3 of a stop difference.

So, here is the question. What do I believe: the filter factor, or the light coming through the filter? Why is there such a huge difference?

Some meters are more sensitive to some colours then others, some films are more sensitive to some colours then others as well. Best way to find out is the old fashioned way:

Take a roll of the film you use most often. Take a test exposure, no filter, put the camera in manual mode, put the filter on. Take one stop under exposed, same shutter speed, one stop over, two stops over, 3 stops over, if you want to be exact do 1/2 or 1/3 stops. Process the film, print the filter less image to get it right, now print the filter exposures using the same time and aperture, develop for the same time and temperature. Now compare the prints, which ever one is closest to the first print, is your own personal filter factor. If the filter factor matches the meter, all is good, if it doesn't then meter without the filter and set the exposure manually.

 

[BetterSense]

I had this problem this morning. I would like to use 400 speed film in my TLR but I don't have the shutter speeds to shoot it in daylight. I was considering using my tiffen 15 deep yellow filter, but I couldn't find a filter factor for it anywhere. If it is 2 stops that would be ideal because I can just squeak by in daylight with 100 speed film.

 

[Chazzy]

The nice thing about standards is that everyone can have their own. This seems to be the approach of those folks making filters since there are so many different naming systems. I also find filter factors mystifying. I have in my hand right now a b+w orange filter (040 - or perhaps you know it by some other number or letter) and written on the ring is "4x" which I am guessing is the filter factor. This means that I open up by two stops. But if I use my spotmeter on a neutral subject to check this, with the filter right up against the spotmeter, I get only 1 stop difference. I have a Hoya light yellow Y (or K2) filter which supposedly has a filter factor of 2, or 1 stop, yet checking with the meter, the needle moves only 1/3 stop. Finally, I have a Kenko YG filter (which apparently doesn't exist because I can find no reference to it anywhere). It is light green and when I bought it (second hand), someone had written on the box "FF=5." But again, with the meter, I get only 2/3 of a stop difference.

So, here is the question. What do I believe: the filter factor, or the light coming through the filter? Why is there such a huge difference?

I'm as frustrated as you are about the different naming conventions. The filter manufacturers should really get together on this. Also, I think that the basic daylight filter factor should be marked on every filter. I realize that filter factors are debatable, depending on circumstances of use, but something would be better than nothing as a starting point, and my memory isn't good enough to remember how many stops to allow for the different filters.

 

[CBG]

Filter Factors for some common films

.........................................Daylight.........Tungsten
No. 8 (yellow)......................2x...................1.5x
No. 11 (yellowish green)........4x...................4x
No. 12 (Deep Yellpw).............2x..................1.5x
No. 15 (deep yellow)............2.5x.................2x
No. 23 (Lt Red)....................6x.......................3x
No. 25 (red)[a]....................8x...................5x
No. 29 (Deep Red)................20x..................10x
No. 29 (red)[a]...............16x..................10x
No. 47 (blue)[a]...................6x..................10x
No. 58 (green)[a].................8x...................8x
No. 61 (Deep Green)..............12x.................12x
Polarizing Filter....................2.5x.................2.5x

 

[RalphLambrecht]

The nice thing about standards is that everyone can have their own. This seems to be the approach of those folks making filters since there are so many different naming systems. I also find filter factors mystifying. I have in my hand right now a b+w orange filter (040 - or perhaps you know it by some other number or letter) and written on the ring is "4x" which I am guessing is the filter factor. This means that I open up by two stops. But if I use my spotmeter on a neutral subject to check this, with the filter right up against the spotmeter, I get only 1 stop difference. I have a Hoya light yellow Y (or K2) filter which supposedly has a filter factor of 2, or 1 stop, yet checking with the meter, the needle moves only 1/3 stop. Finally, I have a Kenko YG filter (which apparently doesn't exist because I can find no reference to it anywhere). It is light green and when I bought it (second hand), someone had written on the box "FF=5." But again, with the meter, I get only 2/3 of a stop difference.

So, here is the question. What do I believe: the filter factor, or the light coming through the filter? Why is there such a huge difference?

Don

Doing your own filter testing is rather complex. I would trust the filter manufacturer's recommendation for daylight and tungsten film and overexpose when in doubt. The lightmeter (or TTL) cannot be trusted at all, because their spectral sensitivity does not match the film close enough.

Always meter without the filter and then compensate with the filter factor!

 

[fschifano]

Amen to that. While one often can get a decent enough exposure by using TTL metering through a filter, it is often inaccurate. To top it all off, films vary in their spectral sensitivity so that even a general recommendation isn't enough. Best to go with what the manufacturer recommends if you have the data. Kodak publishes filter factors for their products and so do Ilford, Fuji, and some others.

 

[Arvee]

+1 on Ralph's spot on advice. After some bad experiences and ruined film, I now believe the mfrs. recommendation for filter factors. I did find, however, that ND filter factors are close to meter indications, reason being that spectral response issues are minimized.

-F.

 

[MattKing]

Filter Factors for some common films

.........................................Daylight.........Tungsten
No. 8 (yellow)......................2x...................1.5x
No. 11 (yellowish green)........4x...................4x
No. 12 (Deep Yellpw).............2x..................1.5x
No. 15 (deep yellow)............2.5x.................2x
No. 23 (Lt Red)....................6x.......................3x
No. 25 (red)[a]....................8x...................5x
No. 29 (Deep Red)................20x..................10x
No. 29 (red)[a]...............16x..................10x
No. 47 (blue)[a]...................6x..................10x
No. 58 (green)[a].................8x...................8x
No. 61 (Deep Green)..............12x.................12x
Polarizing Filter....................2.5x.................2.5x

Thanks CBG - here is a pdf version of the table with the columns aligned and a rough translation of the filter factors into stops:

Matt

 

[dancqu]

Exposure Discussion is the proper place for
this thread. Dan

 

[RJS]

This seems a very vexed question. I have read that metering through the filter is good (written by some good photographers) I have also read Ralph Lambrecht. I've done it both ways and concluded that bracketing is best. Depending on the results desired, i.e. a really dark sky, or a lighter lemon, one or the other exposure might be more satisfactory. If they differ.

And dancqu: don't be such a stickler.

 

[removed account4]

Don

Doing your own filter testing is rather complex. I would trust the filter manufacturer's recommendation for daylight and tungsten film and overexpose when in doubt. The lightmeter (or TTL) cannot be trusted at all, because their spectral sensitivity does not match the film close enough.

Always meter without the filter and then compensate with the filter factor!

thank you !!!

 

[Doremus Scudder]

Here's a vote for metering through the filter.

I've posted on this before in greater detail, but here's the short version:

If you wish to compare tonalities of small areas with and without the filter, metering through a filter with a spotmeter is the only way to get any information at all without shooting and developing film.

Yes, the film and the meter have different spectral responses (even the Zone VI modified meter is not perfect), but these are often negligible and/or able to be compensated for.

Ralph, I found testing to be rather straightforward: simply take your film(s) of choice and do a bracket set with a gray card in sunlight, open shade, and, if you shoot indoors a lot, tungsten light (fluorescents are always a crap shoot). Develop and note the adjustments you need to make for exposing through different meters (e.g., for me, #25/Tri-X320 needs 2/3 stop more exposure than the meter reading). Refine this by doing a standard Zone test with and without the filter to find any development factor (e.g., #25 and Tri-X320 needs -1 to reach desired development, in other words, if you want N, develop N-1). Usually this is only needed for stronger filters. Refine this with field testing.

Then just apply your "fudge factors" when shooting. Yes, this is more fiddly than just applying a manufacturer's filter factor, but I believe it is more accurate and has the advantage of allowing one use metering through the filter as a visualization tool. Of course, if I were shooting 35mm with a TTL meter, I would just bracket...

Don, you might want to check your meter readings under different light sources. You don't mention what you were using when you did your experiment, but I'd be willing to speculate that you were sitting around inside and using tungsten light. Tungsten light is much more heavily weighted toward the red end of the spectrum and will need smaller adjustments for red, orange, and yellow filters and more adjustment for green and blue filters. Also, as I mentioned above, my meter/film combination tends to underexpose by almost a stop when reading through the #25 filter. I adjust for this.

Best,

Doremus Scudder
www.DoremusScudder.com

 

[nicefor88]

I hate filters. They never make a photograph better. The attention gets distracted from the subject to the effect they are supposed to create.

 

[RalphLambrecht]

This seems a very vexed question. I have read that metering through the filter is good (written by some good photographers) I have also read Ralph Lambrecht. I've done it both ways and concluded that bracketing is best. Depending on the results desired, i.e. a really dark sky, or a lighter lemon, one or the other exposure might be more satisfactory. If they differ.

And dancqu: don't be such a stickler.

Yes, I would agree that bracketing around the filter manufacturers recommendation is the best way to get a proper exposure.

 

[RalphLambrecht]

Here's a vote for metering through the filter.

Yes, the film and the meter have different spectral responses (even the Zone VI modified meter is not perfect), but these are often negligible and/or able to be compensated for.

Ralph, I found testing to be rather straightforward: simply take your film(s) of choice and do a bracket set with a gray card in sunlight, open shade, and, if you shoot indoors a lot, tungsten light (fluorescents are always a crap shoot).

Of course, if I were shooting 35mm with a TTL meter, I would just bracket...

Doremus

I would argue that the spectral response differences between film and lightmeters are significant when it comes to filtered light. I made the mistake to buy a set of small filters (Y8, G11, R25) that fit into my Pentax meter. The Y8 results were OK, the G11 results were bad and the result with the R25 right out terrible. Metering without the filter and applying the filter factors was much better! I tried this with a modified and a non-modified Pentax meter, not much difference, save your money!

As far as testing goes, I was mainly talking about B&W contrast-enhancing filters, as in Y8, G11 or R25. Testing the filter factor with a gray card is nice and good, for a GRAY subject. But what is the filter factor of a G11 in the woods? What does a Y8 do for the skin tone when talking a portrait? And how much more light do I need with the R25 in Monument Valley before I lose detail in the foreground rocks (never mind the dark sky then).

35mm or else, the best way is to go with, or bracket around, the recommended filter factor, and in my experience, metering through the filter is just as good as a lucky guess.

By the way, the advertising claim for the modified Zone-VI meter was that 'it closely matches the spectral sensitivity of the human eye'. Well, that doesn't do us any good. It's the film that captures the image, not the retina.

 

[CBG]

I hate filters. They never make a photograph better. The attention gets distracted from the subject to the effect they are supposed to create.

I don't know about "hating" filters, but used inappropriately, sure.

Regardless, there are plenty of times filtration makes sense. Problem is that many times a too strong filter is used and results in an "over the top" effect. Rarely as good as one might want.

Much of the best use of filters is "invisible" and just makes the scene look the way we perceive it, rather than transforming it into something radically different.

 

[CBG]

...a pdf version of the table ...

Much neater! Thanks

 

[Doremus Scudder]

Ralph,

I think we are really saying the same thing.

I advocate reading through the meter after doing tests and finding exposure and development factors. This is more complicated, but more precise than just applying the manufacturer's filter factor and has a couple of advantages.

First, it is coupled to your meter and film combination, therefore bound to be a little more accurate than the generic factor. The mismatch in spectral response between meter and film has to be compensated for, however. That is the reason for the testing. I've found that doing the tests with a grey card gets me much further into the ballpark, so to speak, than just applying the manufacturer's factors.

Second, and most importantly, metering through the filter allows one to compare different areas of a scene and visualize the tonality that will appear in the print. To use your examples, the #11 filter in the woods, when metered through, would show that the green foliage would be rendered lighter than, say, the blue sky. Maybe that's what is desired, or maybe we might choose another filter after determining that. The thing is, the only way to get that information is by taking a meter reading through the filter.


It is likely that my method, which is really a combination of metering through the filter and appying factors, is not totally accurate. However, I have found it closer than just using published factors and find the ability to evaluate portions of the scene with my spot meter while reading through different filters to be very helpful.

Best,

Doremus Scudder
www.DoremusScudder.com

 

[Thomas Bertilsson]

If you meter through the filter, have you taken into consideration the reflected light that comes off the surface of the filter?

 

[keithwms]

To do this right, with full confidence, you'd have to take into account the sensitivity spectrum of the meter, the transmission spectrum of the hot mirror over the meter's cell, the transmission spectrum of the filter, the sensitivity spectrum of the film, and finally... the spectrum of the light actually being used and the reflectance of the subject! These things need to be superposed, and then all will be clear.

Among these factors, what is important and is usually forgotten is that the cell in the meter has a sensitivity curve something like this...

http://www.cs.brown.edu/people/tld/courses/cs148/02/images/wavelength.gif

...and the deep red / IR stuff is cut out by a hot mirror over the cell. So there's no guarantee that the cell's response looks at all like your film's response curve!

Everyone develops his/her own strategy, but what I think is most important is to get a feeling for how much 'slop' there is in the estimation of the filter factor.... and bracket accordingly. There are plenty of popular films with unusual sensitivity spectra i.e. not equally weighted across the visible spectrum.