color temperature

[John Sprung]

why do some people say that 5600K is white light?

 

Not everybody knows as much as you've learned here.

 

 

 

 

-- J.S.

[Jim Nelson]

what do you mean?

 

So that means that 3200K is orange and 5600K is blue and NOT white light?

[John Sprung]

There is no single color temperature that is in any absolute sense the one and only white. Anybody who says that "XXXX Kelvin is white" just doesn't know what they're talking about, or they're assuming you know the context.

 

You get to choose what you want to use for white. If you're shooting film, you have two choices, 3200 or 5600. From there you can hang filters if you want something different. With conventional video cameras, you just push the white balance button and let the camera adapt to whatever light you have (within reason).

 

You can reasonably say "for this scene that we're shooting right now, 5600 Kelvin is white". Later in the day maybe you go inside and for the next scene 3200 Kelvin becomes the new white. Neither one is the only possible white.

 

 

 

 

 

-- J.S.

[Richard Andrewski]

No, flourescents have a green spike because they're mercury arc lamps with a phosphor coating inside. Mercury has a strong emission line at 546.1 nanometers, which is a nice middle green. Except it's not so nice when it shows up on film. Mercury has much stronger lines in the ultraviolet, which is what drives the phosphors to flouresce. The others don't have a green spike.

 

Peak sensitivity for our eyes is 550 - 555 nanometers, but it's not a steep spike.

 

Its not just about the steepness of it, its also about the broadness of it. Our eyes have red, blue and green sensitivity. The green sensitivity is more broader than blue but not as broad as red. Actually the peak sensitivity in where the red and green converge which is more like yellow:

 

Spectral Sensitivity of the Eye

 

What you said is the science behind why a fluorescent tube works, I have an article on my website about that here for more detail:

 

Green Spike or Why Do Fluorescents Have Mercury In Them

 

 

Here's a pretty good question though: is lumen output something absolute and not related at all to human perception or is it all directly related to what our eyes can process? A couple of related questions follow. Is lumen output only related to wattage applied or can the wavelength makeup of the light affect that? What wavelengths of light can we see, what can't we see or see well?

 

As I was trying to say in my first post, there are other subtleties behind all this that most people don't think about. Visible light and what does it mean. You can skip to the last paragraph in this posting if you're not interested in light science.

 

Sure mercury was chosen for its ability to emit UV when excited. Xenon and krypton emit UV too in similar circumstances but have spikes elsewhere than green. Those could be used but why weren't they? For instance, the idea of pulsed discharge xenon fluorescent lamps is a good one but its also acknowledged to not have the lumen output of mercury based lamps. Xenon has a strong spike in the blue and good UV emission. A cool idea but won't go anywhere IMHO because of the efficiency issues. Blue is about the worst color that our eyes can process. Digital camera sensors aren't too crazy about the blue component of light either. I'm not sure about how film feels about blue--probably not as bad as in the digital arena.

 

Metal halide bulbs (including HMI) have the green spike as well because they use mercury and are considered very efficient. Luckily we can make something approaching full spectrum with them. Sodium vapor might actually be considered more efficient but why? It has a strong yellow spike (remember where the red band meets the green one--that's yellow). But, you can't make a full spectrum light or anything even approaching a simulation of white light with it so that's why sodium vapor has a very limited use for those situations where its not a big deal to have a very monochromatic light.

 

In addition, most LEDs in the daylight range of 5000K to 6500K have a strong green spike too but no mercury in them. I think you'll find this to be the case because it boosts the lumen output immensely.

 

 

All this is another way to say that lumen output is a very human perception thing and is therefore not just about the current or wattage being applied but also about what wavelength components are in a light as well to make up the light its emitting. A good illustration: if I have a strong deep blue bulb burning at 2000w and a 1000w pure white bulb, which will be a stronger output in lumens and which will be perceived to be stronger? Because of the wavelengths our eyes are sensitive too obviously the blue bulb is at a serious handicap in both output and perceived efficiency.

[Jim Nelson]

Thank you very much everybody. I finally understood this. It's been bugging me for a long time and now thanks to all of you, it's all clear. Thanks a million.

[Richard Andrewski]

One more thing though, my cinematography teacher said that 3200K is orange and 5600K is blue. why do some people say that 5600K is white light?

 

Of course now I understand that when you white balance to 5600K it'll look white. But if you don't. isn't 5600K blue?

 

 

Thanks

 

He's speaking figuratively not literally--at least I hope so. In any case, you can't get into questioning a college professor. But, at one end the light has a slight red/orange shift. The other end it has a slight blue shift. They're both what we can think of as white light though. Whether he knew it or not, that's what he really meant... ;-)

[David Mullen ASC]

"At around 5000K all wavelengths of visible light are there so the object is "white hot""

 

 

One more thing though, my cinematography teacher said that 3200K is orange and 5600K is blue. why do some people say that 5600K is white light?

 

Of course now I understand that when you white balance to 5600K it'll look white. But if you don't. isn't 5600K blue?

 

 

Thanks

 

To repeat... "white" is relative to what color temp you are balanced to. If your eyes and your film stock are balanced for daylight, then 5500K is "white" and tungsten looks orange. If your eyes and your film stock are balanced for tungsten, then 3200K is "white" and daylight looks blue.

 

5600K is bluer than 3200K (or 3200K is more orange depending on how you want to think of it) but either can be "white" depending on what you pick as the color temp that equals white. For example, movie theater projectors and TV screens are around daylight in color temp, so whites on those screens are also around daylight in color temp.

[John Sprung]

There's a lot of misinformation here, and I don't have much time.

 

For starters, even a very narrow yellow-green LED has a band maybe 25 nanometers wide, like say 575 - 600. Most are a bunch wider. The green spike of mercury is very much narrower, less than a tenth of a nanometer. It arises from electrons dropping energy in discrete quanta, so it may be narrower than can be measured, I don't know. Anyhow, there it is at 546.1 nm. Sodium has a close pair of spikes at 589.0 and 589.6, IIRC.

 

Bottom line, these quantum spikes are a very different thing than other broader light sources.

 

Lumen output is related to human perception, by definition. It's weighted by the CIE 1931 photopic luminosity function, with a maximum of 683 lumens per Watt at 555 nm, which is green. It's not yellow, it's not 600 nm. The scotopic peak is at 550 nm, but that's not so important to us.

 

Could you boost the lumen output by adding green? Sure, but you'd have green light. It wouldn't look white.

 

 

 

-- J.S.

[Jim Hyslop]

To repeat... "white" is relative to what color temp you are balanced to. If your eyes and your film stock are balanced for daylight ...

 

This raises an important point which has not been mentioned yet in this thread (unless I missed it). Your eyes will automatically adjust so that the dominant light source appears white - even if the light is a highly saturated colour. I had a striking demonstration of this when I saw a play lit by a student lighting designer. The designer used a lot of very saturated lavender gels through the first act. At intermission, I looked at the lobby display, which had black and white photographs. My brain had adjusted its "white balance" so that the lavender lights were considered "white" - and thus all the photographs looked green.

[David Mullen ASC]

This raises an important point which has not been mentioned yet in this thread (unless I missed it). Your eyes will automatically adjust so that the dominant light source appears white - even if the light is a highly saturated colour. I had a striking demonstration of this when I saw a play lit by a student lighting designer. The designer used a lot of very saturated lavender gels through the first act. At intermission, I looked at the lobby display, which had black and white photographs. My brain had adjusted its "white balance" so that the lavender lights were considered "white" - and thus all the photographs looked green.

 

Look at your TV screen -- it's even bluer than 5600K, but we don't perceive everything is looking blue on it. Look at a movie theater screen -- that xenon projector bulb is bluer than 5600K as well.

[DS Williams]

Look at your TV screen -- it's even bluer than 5600K, but we don't perceive everything is looking blue on it. Look at a movie theater screen -- that xenon projector bulb is bluer than 5600K as well.

 

 

Yes, it's suprising how well your eyes can adjust to percieve a certain temperature as white, color nuetral light.

 

What I want to know is, what are the 'white balancing limits' of the eye. At what point does your eye stop adjusting? 6500k? 8000?

 

Is there an average, default color balance of the eye?

 

I mean certainly, light with even spectral distribution, like that of a brute arc of HMI, looks white at first glace, whereas tungsten lights require me to adjust more...at least that's me.

[Chris Keth]

Yes, it's suprising how well your eyes can adjust to percieve a certain temperature as white, color nuetral light.

 

What I want to know is, what are the 'white balancing limits' of the eye. At what point does your eye stop adjusting? 6500k? 8000?

 

Is there an average, default color balance of the eye?

 

I mean certainly, light with even spectral distribution, like that of a brute arc of HMI, looks white at first glace, whereas tungsten lights require me to adjust more...at least that's me.

 

Your eyes can adapt to an amazing range of conditions. I don't know their limits but they seem to be well above what is required. Try this: find a strongly colored wall outdoors somewhere. Something painted a nice bright, saturated color. Get close enough that it fills your entire field of vision and stare at it. Keep staring. Give it a few minutes, your eyes won't like it much, but give it a few minutes. Still staring? Turn around and look around. Your vision will be tinted the complementary color to the wall because your eyes started to compensate for all that color.

[John Sprung]

What I want to know is, what are the 'white balancing limits' of the eye. At what point does your eye stop adjusting? 6500k? 8000?

 

That's a tough question. First you have to decide how good is good enough. One way to do that might be to make some color test patches of blues and oranges having a barely noticeable difference between two shades of each color, when viewed at, say, 4000 Kelvin. Then look at them under light of variable CT, and run the CT up and down until you reach a point where those differences go away. Get a few hundred volunteers to do that same test, and you might come up with some useful data. You might also be able to get a PhD for doing it. (BTW, I looked at my previous guess on this, 2600 - 6500, and I'm pretty sure I was way wrong on the high end. New guess, probably more like 2600 - 10,000 Kelvin.)

 

Is there an average, default color balance of the eye?

 

My guess is probably not. We adapt so quickly that even if there were one, it might not matter much. One possible test would be to spend some time in total darkness, perhaps an hour or more, then go to a tungsten environment. Do it again, only go to a daylight environment, with the same intensity. How long does it take to adjust? Try it with other CT's to see if there's one that doesn't need adjustment. Also, does it matter what CT you were looking at just before you went into the total darkness? Does it matter where you grew up? Maybe people from overcast/cloudy environments will have more range on the high end?

 

Another thing to look at is the ability to adapt to things that are a little off the CT curve, like flourescents.

 

 

 

 

-- J.S.

[John Sprung]

Your vision will be tinted the complementary color to the wall because your eyes started to compensate for all that color.

 

Part of it's in the brain, and part is chemical, in the retina. It's the chemical part that re-adjusts somewhat slowly, producing this effect.

 

 

 

 

-- J.S.

[Jim Nelson]

Thank you very very much everybody :)

[Jim Nelson]

may I ask one last question please.

 

Which type of film (daylight or tungsten) do you use when you have lights that look to the eyes: green, red, magenta, orange etc? The lights I am referring to are the ones you can see at night clubs or at concerts. And I want them to look the way I see them with my eyes.

 

 

Thanks :)

[Dimitrios Koukas]

may I ask one last question please.

 

Which type of film (daylight or tungsten) do you use when you have lights that look to the eyes: green, red, magenta, orange etc? The lights I am referring to are the ones you can see at night clubs or at concerts. And I want them to look the way I see them with my eyes.

 

 

Thanks :)

 

That's a difficult question, because now in stages we use different types of light bulbs, some of them are tungsten (parcans) some other MSR's (color temperature: 7200 K) and many more like HMI's, LED's, aircraft lights etc.

So if you want to make all of this look excactly like you see it, you have to correct ''the source'' [what is behind the colour gels or dichroic heads] to the desired Kelvins. 3200-5600-4300 or whatever your stock or your white balance is.

Example : if you decide that you are filming with a 3200 K stock, you have to correct the MSR's and HMI's to 3200 K and leave the par cans uncorrected.

There other factors that come to this game though, like loosing too much output from the fixtures from correcting them.

Dim

[Tom Jensen]

may I ask one last question please.

 

Which type of film (daylight or tungsten) do you use when you have lights that look to the eyes: green, red, magenta, orange etc? The lights I am referring to are the ones you can see at night clubs or at concerts. And I want them to look the way I see them with my eyes.

 

 

Thanks :)

 

You will more than likely use Tungsten film to keep it on the cool side. Nightclub lights are usually background, accent lights. If you are lighting a face you will more than likely have tungsten lights to do that. Keep it simple, use Daylight film in the daylight.

[Jim Nelson]

what I meant is that I want those lights to look green, red, magenta, orange etc. Because that's how I see them with my eyes. I don't want to make them look white.

 

Also if I use the wrong film, will the magenta light look more magenta on the film, will the green light look more green etc?

 

 

thanks

[Hal Smith]

That's a difficult question, because now in stages we use different types of light bulbs, some of them are tungsten (parcans) some other MSR's (color temperature: 7200 K) and many more like HMI's, LED's, aircraft lights etc.

So if you want to make all of this look excactly like you see it, you have to correct ''the source'' [what is behind the colour gels or dichroic heads] to the desired Kelvins. 3200-5600-4300 or whatever your stock or your white balance is.

Example : if you decide that you are filming with a 3200 K stock, you have to correct the MSR's and HMI's to 3200 K and leave the par cans uncorrected.

There other factors that come to this game though, like loosing too much output from the fixtures from correcting them.

Dim

 

Some intelligent lighting has CTO as one of the options in their color filter wheels for that very reason. An example of one that does is my High End Cyberlight Lithos.

[David Mullen ASC]

what I meant is that I want those lights to look green, red, magenta, orange etc. Because that's how I see them with my eyes. I don't want to make them look white.

 

Also if I use the wrong film, will the magenta light look more magenta on the film, will the green light look more green etc?

 

 

thanks

 

You can't really choose wrongly when the colors are that saturated and intense. Take a DSLR into a night club and take some photos in tungsten and daylight mode and you'll see what I mean -- a red gelled lamp is going to be deep red either way. The main difference will be the blueness of the MAC projectors and xenon follow-spots, which are daylight-balanced.

 

Most people use 500T stock because they need the 500 ASA speed, not so much because they need tungsten-balance. Though most stage PARCAN's are gelled tungsten lamps so it's a good starting point.

[Dimitrios Koukas]

@Sam Chandima

 

Just don't excpect to see black lights (UV) recorded as you see them...

;)

[Walter Graff]

I see a lot of confusion here in this topic. Perhaps this is because the notion of color temperature has it's roots in a science distant from lighting for film and TV and the translation isn't 100%. It's a pretty simple notion but I think sometimes we over analyze stuff. And many times web searches don't help us as much as confuse us. This is an article from some time ago that I normally reference for astronomy talks I give, but it will explain a lot to the masses here too. Do read the comments too. It will show you just how confusing the topic is when one over thinks it.

 

http://blogs.discovermagazine.com/badastro...no-green-stars/

[Jim Nelson]

"You can't really choose wrongly when the colors are that saturated and intense. Take a DSLR into a night club and take some photos in tungsten and daylight mode and you'll see what I mean -- a red gelled lamp is going to be deep red either way."

 

But the thing is, I took a photo of an orange light. And when I balanced to tungsten, it looked exactly how I see it with my eyes, in other words orange. But when I balanced to daylight, it looks way too orange and not too realistic.

 

Is it that whenever you have artificial lights (no matter what their color: blue, magenta, green, orange etc like the lights at night clubs), then you balance to tungsten? And when you have natural light (the sun) then you balance to daylight?

Edited April 18, 2009 by Sam Chandima

[DS Williams]

Excellent point Walter. Color temperature is fairly simple in concept, but we over think it too much. I think much of this confusion comes from the inherent differences between film stocks and our human visual system as a whole (retina, brain, ect)

 

Great article as well, I suggest everyone read it. Take these concepts outside of cinematography for a change.

Edited April 18, 2009 by DS Williams