5 replies to this topic

### #1 Malik Sajid

Malik Sajid
• Basic Members
• 130 posts
• Student
• Lahore, Pakistan

Posted 11 December 2008 - 10:35 AM

As you know, i am a student learning cinematography, i have to ask such basic questions.

Is color temperature depends on the source or the bulb inside? suppose we have one source, and on the other hand we have multiple units of the same light. Does color temperature remain the same, or it will be higher with multiple sources?
• 0

• Sustaining Members
• 7117 posts
• Cinematographer
• Los Angeles, Ca

Posted 11 December 2008 - 10:57 AM

Color temp is based on the bulb which is within a source. If you had something like a 9-light, well then ll the bulbs inside of it would be 3200K, as they would in all 9-lights. If you had 3000000000000000 or so 1Ks (wouldn't that be nice!) they'd all have the same color temp.
The only time this is different is with HMIs, whose bulbs will change color temp as they age, so sometimes it becomes important to match them all together, using a color meter and color correction gels on the lights. Personally I've never had to do this as often they're "close enough," but it's important to be mindful of.
Don't forget too that color temperature isn't as important as the Mired Shift. http://www.cinematog...alculations.htm
In essence thing of it like this, the difference in color between 2900K and 3200K is a lot greater than between say 5600K and 6000K. To account for this, there is the Mired scale which has some math with it.
• 0

### #3 David Regan

David Regan
• Sustaining Members
• 218 posts
• 1st Assistant Camera
• New York

Posted 11 December 2008 - 11:01 AM

It can be defined, more or less if I remember correctly, as being derived from the color a black body radiator would turn when heated. As this is heated, it turns from black to red, to yellow, to blue. Thus the higher the temperature, the cooler it is. So 3200K tungsten lamps have the equivalent color output that we would see if we heated this black body radiator to 3200 degrees Kelvin. Likewise if we heated it to 5600 degrees kelvin, it would be theoretically the approx. color of typical daylight sources.

As for the second part of your question, color temperature remains the same (theoretically) regardless of the number of sources. If you throw up 30 tungsten units and aim them at something, and then turn 29 of them off, you should still have the same color temp.

Bear in mind, color temperature is not always exact. In tungsten units variations in voltage can alter the exact color output of a light. Likewise HMI and Kinoflo bulbs can deviate from 5600K, which can change as the bulbs age.
• 0

### #4 David Mullen ASC

David Mullen ASC
• Sustaining Members
• 19760 posts
• Cinematographer
• Los Angeles

Posted 11 December 2008 - 11:01 AM

Well, in this case (color temp) the bulb IS the source generally, so to speak, and it produces visible light made up of different wavelengths - a bias towards one end of the spectrum compared to the other is measured by its color temperature in Kelvin. Visible light is bordered at one end by infrared wavelengths and the other end by ultravoilet wavelengths -- within those two ends is what our eyes can see, red to deep blue.

Now, this assumes a "full spectrum" source containing all the wavelengths, but some sources have a partial spectrum with gaps, like some industrial gas discharge lights (sodium streetlamps, metal halide lamps, etc.) These may have a bias towards one end of the spectrum or the other, a color temp, but the missing wavelengths also produce a color spike, usually towards the green. For example, a Cool White fluorescent tube often has a color temp around 4800K (slightly warmer than daylight) but has a green spike as well.

Multiple sources of the same color temp, let's say, 3200K, don't change the color temp just by adding one light next to another, it just changes the overall output/intensity. Sources mixed of different color temps, especially if blended in some manner (like passing through a diffusion frame) will cause some averaging of their colors, depending on the intensities of the different sources. For example, a 575w HMI (5600K maybe) and a 2K tungsten (3200K) shone through a diffusion frame may produce a soft light that is inbetween 3200K and 5600K... but a 650w tungsten Tweenie put next to an 18K HMI is hardly going to shift the color temp because the output is so dominated by the 18K HMI.

What you bounce or diffuse the light with may impart some shift in color temp. Aging bulbs, voltage drops, etc. may also affect the color temp of the light.

http://en.wikipedia....lor_temperature
http://en.wikipedia....i/Visible_light
• 0

### #5 Malik Sajid

Malik Sajid
• Basic Members
• 130 posts
• Student
• Lahore, Pakistan

Posted 13 December 2008 - 07:30 AM

Arrright...thanks Adrian, Regan and David......i really appreciate.....i guess i have got something about color temperature now.

So can we say that color temperature determines the color?
• 0

### #6 David Mullen ASC

David Mullen ASC
• Sustaining Members
• 19760 posts
• Cinematographer
• Los Angeles

Posted 13 December 2008 - 11:14 AM

So can we say that color temperature determines the color?

In the warm to cold axis, it doesn't tell you about hues like green, purple, yellow, magenta, etc. Two different colors, like Apricot and Flame Red, perhaps will have similar color temp effects.

Just look at camera filters that can correct 5600K to 3200K, like a Decamired, a Coral, a 85B... they may all cause the same MIRED shift, but they don't match each other in color, some are yellower or more magenta than others.
• 0