posted 12-07-2000 02:05 PM
Suppose you had a light bulb whose filament would never melt, no matter how hot it got, and that would take as much electric current as you cared to put through it. With no current flowing through it the filament is at room temperature and emits no visible light. Now if you start to apply electric current to it, slowly raising the temperature the filament will start to glow. Experience tells us that it will first glow red, but as we raise the temperature the light will slowly turn from red to white.As we perform this experiment let's pass some of the light through a prism, splitting the light into its various colors, and projecting those colors on a screen. At first the screen is blank. Then as we apply current to the filament we will see red at one end of the spectrum. As increase the current we will see green, then blue and finally violet appear, but they will not be as strong as the red. However, as we continue to apply more and more voltage the other colors grow stronger until the green is as strong as the red, then the blue and finally all the colors are of the same intensity and the filament glows white hot.
Sunlight exhibits the characteristic that all the colors in the visible spectrum are equally intense, provided that the light has not been tainted by passage through the atmosphere. Light from most man made sources is not equally distributed across the visible spectrum but is more intense toward the red end.
So what is color temperature? Well it is not the temperature of your light bulb. When scientists do the above experiment they do not use a filament. They use a hollow block of metal with a hole drilled in it and they measure only the light that comes from inside the block. This is called a "cavity" experiment and color temperature refers to the temperature of the cavity.
Daylight film has its dyes adjusted for a light source of 5500°K. With this light the intensity of all the colors is pretty much equal. Household tungsten filament bulbs have a color temperature of about 2800°K and the red is stronger. Photos taken with daylight film under tungsten light will produce Caucasian skin tone with a decidedly reddish or orange tint. To correct this either use tungsten balanced film (which is less sensitive to red) or use a blue filter to attenuate the red. Or use electronic flash. The color temp of electronic flash is 6000°K, close enough to 5500°K for most people's tastes.
Video cameras have some provision for adjusting the color temperature. Most consumer video cameras (the only kind I have ever owned) do this automatically. More expensive pro cameras may give you more control.
If you are rich enough to afford a color temperature meter and a complete set of filters you can shoot whatever film you want and adjust the color temp to your liking. If you are poor like me just buy an 80A filter and use it when shooting under tungsten. When shooting under fluorescent light use an FL filter. These cost less that $20 each.
Finally, I want to say to those of you out there with Ph.D.s in physics that I realize that I have oversimplified things a bit.