Color temperature is a simplified way to characterize the spectral properties of a light source. While in reality the color of light is determined by how much each point on the spectral curve contributes to it's output, the result can still be summarized on a linear scale..

This value is useful eg. for determining the correct film in photography depending on the lighting (resp. for determining the white balance in digital photography), and for specifying the right light source types in architectural lighting design. Note, however, that light sources of the same color (metamers) can vary widely in the quality of light emitted. One may have a continuous spectrum, while the other just emits light in a few narrow bands of the spectrum. A useful way to determine the quality of a light source is it's color rendering index..

Low color temperature implies warmer (more yellow/red) light while high color temperature implies a colder (more blue) light. Daylight has a rather low color temperature near dawn, and a higher one during the day. Therefore it can be useful to install an electrical lighting system that can supply cooler light to supplement daylight when needed, and fill in with warmer light at night. This also correlates with human feelings towards the warm colors of light coming from candles or an open fireplace at night..

Standard unit for color temperature is Kelvin (K)..

(The kelvin unit is the basis of all temperature measurement, starting with 0 K (= -273.16 C) at the absolute zero temperature. The "size" of one kelvin is the same as that of one degree Celsius, and is defined as the fraction 1/273.16 of the thermodynamic temperature of the triple point of water, which positions 0 Celsius at 273.16 K.).

Technically, color temperature refers to the temperature to which one would have to heat a theoretical "black body" source to produce light of the same visual color..

Some typical color temperatures are:.

1500 K Candlelight2680 K 40 W incandescent lamp3000 K 200 W incandescent lamp3200 K Sunrise/sunset3400 K Tungsten lamp3400 K 1 hour from dusk/dawn5000-4500 K Xenon lamp/light arc5500 K Sunny daylight around noon5500-5600 K Electronic photo flash6500-7500 K Overcast sky9000-12000 K Blue sky.

And just so know one gets upset I just copied this from another web site. I would have simply included a link but DP is so freakin picky about links this was easier..

Maddog.

Olympus E-500, Olympus E-510..

Well, I'm new to photography, but dabbling in 3D graphics, I learned a bit about those values. They are called color temperature and are in fact, temperatures in degrees Kelvin (if you need very rough conversion from degrees F divide temp. by half and add 230. ).

When dealing with lightsources that produce light by heating something (like fire, common light bulb, or Sun) the color of light they produce _approximately_ corresponds to their temperature. So, light bulb is 3000-3600 K, bright Sun, - approx 9000, etc....

When dealing with light produced by other means, or one that is somehow filtered, or even incadescent light produced by not-so-uncommon chemicals this "temperature" makes little sense, of course, and should be selected "by eye"..

Again, I'm very new to photography, and may be grossly wrong...

Setting your color temp explicitly in-camera matches the camera's response to the temp of the light source so that colors appear normal..

But you might be confused when setting the temp by degrees in an editing tool like Photoshop... the color shifts opposite to the absolute change in degrees. For example, if your image was shot under tungsten (lightbulb) lighting (without setting the camera for tungsten color temp), the scene will tend to be very orange..

Setting the temp to tungsten temperatures in the editor does not make it *more* orange (it's not setting an aboslute temp value)... it is setting a *compensation* for the indicated temp of the tungsten lightsource, making it more blue to counteract the orange..

That confuses many people first starting to deal with color temperature, but just keep in mind that the editor is compensating in the opposite direction on the color scale to make things look as they should..

Ecables wrote:.

Can anyone give a brief overview of the "K" values when setting WB?They are pretty greek to me, how does a photographer know when tochoose 5200 or 5700?.

..

You got some amazing explanations, here, so this tutorial is probably overkill, but....

Http://www.cambridgeincolour.com/tutorials/white-balance.htm..

Ablack wrote:.

They are called color temperature and arein fact, temperatures in degrees Kelvin (if you need very roughconversion from degrees F divide temp. by half and add 230. ).

Two minor adjustments about the Kelvin scale:.

* there's no such thing as "degrees Kelvin", it's just "Kelvins" like"Watts". Since there is an absolute zero point which you can'tgo below, and not just arbitrary zeros like freezing water, theusage of "degrees" does not make sense..

* the rough conversion is easier in degrees Celsius, since they arethe same (273.15 Kelvins = 0 degrees Celsius, 373.15 Kelvins= 100 degrees Celsius). The ratio from degrees Fahrenheit,you will recall from school, is 5/9, which I guess is close enoughto "half" if you really only need "very rough" conversion.  .

[ e d @ h a l l e yc c ] http://www.halley.cc/pix/..

Nhanks for the reply, it was educational..

Ed Halley wrote:.

* there's no such thing as "degrees Kelvin", it's just "Kelvins" like"Watts". Since there is an absolute zero point which you can'tgo below, and not just arbitrary zeros like freezing water, theusage of "degrees" does not make sense..

Interesting... As much, as I can tell there's no such distinction in my native language (in the papers and over the net they just use shorthand, and I haven't been talking physics personally for quite a while.).

Is this generally accepted convention, or something only some groups use?.

* the rough conversion is easier in degrees Celsius, since they arethe same (273.15 Kelvins = 0 degrees Celsius, 373.15 Kelvins= 100 degrees Celsius). The ratio from degrees Fahrenheit,you will recall from school, is 5/9, which I guess is close enoughto "half" if you really only need "very rough" conversion.  .

Yes, it's rough, but not quite that badly as it may seem, less than 5.6% off..

If it is really important, then you can add 55 for every thousand, 5.5 for every hundred, and .5 for every ten after the division by 2. This is easier than multiplying by 5/9 and scales nicely..

(often there's no even 5% precision in the measurement you need to convert in the first place. If someone tells me "It was about 10 ft tall" I'd look like ... if i'd took out my calculator. So it's just 3 m. ).

In this particular case, where temperature is only a rough guide anyway, 5% error is less importanrt than ease of use....

And yes, 273-32 is 241, not 230.....