Tell me about color filters

[BetterSense]

I'm a noob photographer and I never gave white balance any thought. But when I use color print film for snapshots and shoot indoors in tungsten light, everything is orangey. Under fluorescent light sometimes it's greenish. I've come to realize that maybe you can cancel these things out with filters, if you were willing to sacrifice the speed, but I've never tried, since I have no filters. What are the most useful color filters and what kind of situations should they be used in, or what kind of effects can you get? I've heard about people that use yellow filters with B&W too, because it supposedly increases contrast.

 

[Bob-D659]

Try here, lots to read and learn.

http://www.schneideroptics.com/info/stillphotography.htm

 

[mpirie]

The use of filters in colour photography tends to be for the correction of colour temperature. Normal "daylight" film is tuned to respond to a colour temperature of around 5500 degrees Kelvin (which equates to midday in Washington on the longest day if my memory serves me right). If you were in this situation, you shouldn't need any correction filters.

Indoor lighting (although it looks "white" to our eyes) is not white as far as the film is concerned. Tungsten filament lamps have a temperature of around 3200 degrees Kelvin, so you can see that there will be a difference of 2300 Kelvin. To correct this situation, you would use a "tungsten to daylight conversion" filter (normally called 80a/b/c). Flourescent indoor lighting has a different temperature again, and yet another filter (FL/Day) would be required. The same applies to mercury and sodium (street) lamps.

Of course, there are (or were) tungsten tuned films which, if you used indoors needed little or no correction, but if you took them outdoors, the resulting blue cast would be horrendous.

Many of the casts on your print film can be partially removed during printing, but it's better to do the correction at the taking stage because the relationships of the other colours is retained. If you try to filter our casts during printing, you affect all the other colours too.......and usually with poor results.

The filters used in B&W photography are used to modify the response of panchromatic (sensitive to most colours) film. The normal filters (yellow, green, orange, red) will cut back on the transmission of their complimentary colours, so example a red filter will cut the amount of blue getting through to the film, thereby underexposing it and making a blue sky darker.

Mike

 

[Ralph Javins]

Good morning, BetterSense;

Two publications recommended are the Kodak book on filters and the books and pamphlets published by Tiffen. Ira Tiffen also provided some of the commentary in the later Kodak books on filters.

 

[nickandre]

You need to convert the tungsten balanced light to daylight. There is more red and less blue and similar green. You have to block both the red and the green down to the blue levels, which means you lose a stop and a third. Your 400 speed film becomes ASA 160. The only tungsten balanced film available is the 64T slide film, which is quite slow.

 

[Puma]

Which KR and KB filters do I really need?
Where can I find a Kodak value to mired value conversion chart?

Thank you,

Puma

 

[Leigh B]

Flourescent indoor lighting has a different temperature

A common mis-conception.

Color temperature is ONLY the ratio of red to blue light. It knows nothing about green.

The problem with fluorescents is that they exhibit excess green as compared with the spectrum of "hot" lights like tungsten or the sun. The only solution to this problem is to use a magenta filter (which is minus green), or a filter specifically designed to correct the excess green.

- Leigh

 

[Lee L]

Color temperature is ONLY the ratio of red to blue light. It knows nothing about green.

This is not an accurate statement. Color temperature is based on the color content of a continuous spectrum (with a relatively smooth curve) emitted from a theoretical 'black body' at a given temperature in Kelvin. This spectrum does contain green light, and the emitted spectrum contains different ratios of short (blue) vs. long (red) wavelengths and all the wavelengths between them as the temperature rises or falls.

One of the problems with fluorescents is that their color spectrum has large spikes of certain colors depending on the phosphors used to coat the tube, often giving them a disproportionately greater green content (a 'spike' in the power spectrum) relative to black body radiation.

See the Wikipedia entry on color temperature, and especially the section on spectral power distribution near the end of the entry.

Modern flourescents are assigned color temperatures that are generally accurate in character, but the phosphor 'spikes' can cause them to be off somewhat from what you'd expect from black body spectral content.

Lee

 

[Leigh B]

Yes, color temperature is by definition the spectrum of a black body heated to the specified temperature.

However, its measurement is accomplished by analyzing the red and blue components of the spectrum with absolutely no regard to green.

That's why color temperature meters like my Gossen Color Pro 3F give two readings, one for the color temperature and one for the green component.

The magnitude of the green component of a black body radiation curve can be inferred. Fluorescent lights do not have an emission spectrum that conforms to the black body model.

I'm quite familiar with spectral distribution, thank you. I used to teach color theory.

- Leigh

 

[2F/2F]

I just made a post in a similar thread: (there was a url link here which no longer exists).

 

[Lee L]

However, its measurement is accomplished by analyzing the red and blue components of the spectrum with absolutely no regard to green

You are conflating what Gossen does specific to metering color balance for photography with the definition of color temperature, and assuming that their method is the only one used. Gossen chooses one way to do it that works well with three layer color film, and their meter does measure green relative to magenta (a mix of both blue and red). Their method isn't the only way to measure color temperature.

However, your statements about measuring color temp with only blue and red, disregarding green are simply wrong, and misleading.

Astronomers measure color temperature with Ultraviolet, Blue, and Visible (green-yellow) passband filters, that don't reach deep into red. And yes, they do use green wavelengths as a component to measure color temperature. That's where the V component filter peaks in sensitivity.

You can also use a spectrograph to measure color temperature, which includes green components.

It's nice that you taught color theory. My mom, who is an artist, taught me additive and subtractive color theory when I was a child. I've also used a Minolta color meter for commercial/advertizing photography on location and in the studio.

I also wrote and taught curriculum for physics labs on color temperature at a top ranked college, which was reviewed and also taught by physicists with Ph.D.s from Cornell and U.C. Santa Barbara.

You should recognize that there is more than one way to measure color temp, and that stating that it's always and only a blue-red ratio is wrong.

Lee

 

[Leigh B]

However, your statements about measuring color temp with only blue and red, disregarding green are simply wrong, and misleading.
Astronomers measure color temperature with Ultraviolet, Blue, and Visible (green-yellow) passband filters, that don't reach deep into red. And yes, they do use green wavelengths as a component to measure color temperature. That's where the V component filter peaks in sensitivity.
I also wrote and taught curriculum for physics labs on color temperature at a top ranked college...
You should recognize that there is more than one way to measure color temp, and that stating that it's always and only a blue-red ratio is wrong.

You'll forgive me for presenting an answer that's relevant specifically to photography.

This is after all a photography forum, not one for astronomers or physicists.

Since you appear to be well-versed in the theory, you should know that the "color temperature" of any visible black-body spectrum can be calculated by comparing the amplitudes at any two points on the curve. Accuracy is enhanced by choosing two points that are as far apart as possible in wavelength (i.e. color).

It's convenient and less expensive from a manufacturing standpoint to measure red and blue components and compare them. This DOES NOT include green, which is measured separately if at all.

We're dealing with visible light here, with CT in the range roughly 2500° to 6500° k.

Photographers don't care about UV.

- Leigh

 

[olleorama]

Reminds me of this:

 

[Lee L]

Photographers don't care about UV.

That must be why no one manufactures or buys UV filters for camera lenses.

Lee

 

[Leigh B]

That must be why no one manufactures or buys UV filters for camera lenses.

Are we to be perpetually bombarded by your nonsensical remarks?

My comment was in the context of color temperature calculation, as a rebuttal to your mention of UV measurements being used for that purpose.

As to UV filters, they're sold and used primarily as physical protection for the front element of the lens, not for their attenuation characteristics.

- Leigh