Metering with colour filters

[Barney]

The generic advice is that increase exposure by "x" stops.

Is there not a better way than this ?

How is this working here, have I got it right, take for example yellow filter if I add a stop I am making yellow and white even brighter when in effect I only want to darken blue. Would the filter not be more effective if I metered off something yellow and then exposed for that without opening up a stop to get maximum effect on the blue sky.

 

[Woodsy]

If you have a spot meter, point it at a bright white cloud and take a reading (or a grey card, etc). Stick your yellow filter in front of the meter and take a reading from the same location.

Adding anything in front of the lens will never make anything brighter in absolute terms, only brighter than something else relatively speaking - which in essence is the whole point.

In order to fully understand how a filter will affect colour, and hence relative tones on B&W, you need to understand the spectral response of the filter (something a good datasheet should show), and to a lesser extent with panchromatic films, the response of the film itself assuming the response is fairly flat. You can do fairly indicative measurements of this again with your spot meter. Find some common blues, yellows, reds, greens, and spot meter with and without your filter in the way.

Colour filters have these almost throw away correction values because for the most part they are fine, especially when taking the exposure latitude of the film into account (generally +/- 1 stop doesn't make a wild difference unless you're using a particular film. Old Adox CHS25 used to be quite unforgiving of over exposure when pitched against most other modern emulsions.

 

[Barney]

If you have a spot meter, point it at a bright white cloud and take a reading (or a grey card, etc). Stick your yellow filter in front of the meter and take a reading from the same location.

Adding anything in front of the lens will never make anything brighter in absolute terms, only brighter than something else relatively speaking - which in essence is the whole point.

In order to fully understand how a filter will affect colour, and hence relative tones on B&W, you need to understand the spectral response of the filter (something a good datasheet should show), and to a lesser extent with panchromatic films, the response of the film itself assuming the response is fairly flat. You can do fairly indicative measurements of this again with your spot meter. Find some common blues, yellows, reds, greens, and spot meter with and without your filter in the way.

Colour filters have these almost throw away correction values because for the most part they are fine, especially when taking the exposure latitude of the film into account (generally +/- 1 stop doesn't make a wild difference unless you're using a particular film. Old Adox CHS25 used to be quite unforgiving of over exposure when pitched against most other modern emulsions.

Thanks for that jonanthan, will test that right away !

 

[sk66]

The blue component of light, which the yellow filter absorbs, is also a component of the light falling elsewhere else. I.e. it reduces the light everywhere, but it is most apparent where the other components of light are already absorbed (blue parts of the scene).

 

[Petrochemist]

The blue component of light, which the yellow filter absorbs, is also a component of the light falling elsewhere else. I.e. it reduces the light everywhere, but it is most apparent where the other components of light are already absorbed (blue parts of the scene).

Usually but not always.
The physiogram below is composed purely of monochromatic green light
laser trace by Mike Kanssen, on Flickr
I also have loads of traffic trail shots that are mostly black, with red, blue, amber & white lights - a yellow filter here would reduce the blue significantly, the white noticeably with little effect on the red & amber. Blue is generally rather unusual in these sorts of shots & red can sometimes be predominant.

The best way is usually via TTL metering, but filter factors are a good alternative for use with normal scenes in everyday whiteish lighting.

Filter factors become problematical when dealing with filters that block specific sharp bands, such as the neodymium filter, which blocks sodium D lines (low pressure sodium lights & some other applications) but has relatively little effect on normal scenes just reducing yellows a bit. I don't find they work well for many of the shots I take with polarisers as the alignment of the filter often makes a huge difference.

They might also be fairly useless if your film is not panchromatic. - A red filter used with ortho film makes little difference.

 

[sk66]

Usually but not always.

Yes; I did assume full spectrum daylight/landscape photography...

 

[Petrochemist]

Yes; I did assume full spectrum daylight/landscape photography...

I wouldn't say 'full spectrum' as that term is used to indicate cameras specially converted to have increased response to IR & UV (around 300-1100nm overall).
Two of my digital cameras are full spectrum converted. Filter factors don#t work for them either as most filters actually transmit considerable near IR and even TTL metering can be misleading for them...

 

[Barney]

Well I am gobsmacked here I just had to check the date of this posting in disbelief.

I received my copy of Ansel Adams forty photos yesterday afternoon and I have just read his MO on Sand Dunes and Sunrise and in the first few paragraphs Ansel describes how Edward Weston advised him not to use a filter factor for yellow a filter!
As it brightens the image too much !

Unbelievable.

 

[sk66]

Well I am gobsmacked here I just had to check the date of this posting in disbelief.

I received my copy of Ansel Adams forty photos yesterday afternoon and I have just read his MO on Sand Dunes and Sunrise and in the first few paragraphs Ansel describes how Edward Weston advised him not to use a filter factor for yellow a filter!
As it brightens the image too much !

Unbelievable.

Makes sense since sand dunes and the sun have a lot of yellow/warm tones that won't be affected by the filter.

 

[zx9]

Makes sense since sand dunes and the sun have a lot of yellow/warm tones that won't be affected by the filter.

A trap I fell into with TTL metering using a yellow filter on a bright sunny day in Oxford, I found that the yellow sandstone buildings were rendered much lighter than I would have liked. I have since stopped habitually using yellow filters just because it is sunny and I want to render the blue sky darker and try to think it through a bit more, a reminder not to be stuck in a routine.