Color Temperature, the Eye, Film and Video

[Steve London]

So if my eye and brain can immediately adjust to different lighting enough that white looks pretty white and skin tones look right under 3200K or 5600K illumination when I'm there in person viewing the scene, why don't I do this same trick when viewing a photograph or electronic image of it?

 

If I stare for a while at a photograph taken with daylight balance but tungsten illumination, will it eventually stop looking so orange and I'll perceive fairly correct colors? If so why, and why isn't it immediate? (I should do the experiment, I suppose.)

 

You get the idea, and thanks for detailed explanations about this psychobiological phenom.

[Sander van de kerkhof]

I'm by no means an expert on this matter so this is more or less a guess, But i like the question and it made me think.

 

You need light to emit a certain color temperature, this light is what is taken in by your eye and what you compensate for in your brain. you can fool this by looking through a piece of CTB for 10 minutes and you will see that for a couple of seconds your brain is not correctly balanced. this works because light comes through the ctb just at a different color temp.

 

The photograph you stare at does not emit light in a certain color temperature so a daylight balanced picture that has been lit with tungsten wil print as a yellow/orange picture and that is wat u see. and because this is accurate your brain does not compensate for it.

You would never be able to see a painting or a movie in accurate color because you would constantly compensate the color in your brain.

 

Like i said just a guess,

[Hal Smith]

So if my eye and brain can immediately adjust to different lighting enough that white looks pretty white and skin tones look right under 3200K or 5600K illumination when I'm there in person viewing the scene, why don't I do this same trick when viewing a photograph or electronic image of it?

The eye integrates the entire visual field to auto white balance. The photo isn't the entire visual field, it's being illuminated by the color temperature of the overall lighting. As a result the photo looks orange or whatever.

 

If you're ever in a color timing situation, be certain to have a white light in the room that's got the color temperature of what you want your final result to match and look at it every so often to rebalance your eye. Otherwise you can get results that on later viewing will look hideous. I think this may be why so many current movies look like cr*p on the screen, someone like a corporate suit, with no sophistication about color vision, was in the timing sessions making idiot decisions. I can just hear that person saying "I want it more blue" not understanding that they've been watching a increasingly blue image and their eye now thinks 9000K is white.

 

The best demonstration I've seen of the eye's auto white balance was at one of the Broadway Lighting Master Classes I attended in NYC. One of the top projection designers, Beverly Emmons, gave a lecture demonstration titled something like "Why Sometimes a Light Doesn't Look Like You Expected'. She set a dress dummy up on stage with a neutral white dress on it (from the Broadway production of "The Heiress" no less). She first lit it from one side at a fairly low level with a medium magenta gel on a ellipsoidal fixture. Leaving that light on, she then lit it from the other side with a somewhat brighter light running the same gel. She asked "which side looks whiter", and the roughly 100 lighting designers in the room all agreed the brighter side looked whiter. She proceed to go from side to side slowly increasing intensity and asking the same question. When she got the dummy's intensity level up to may 100FC or so she pulled her rabbit out of the hat. She hit the dummy from the balcony rail with an equally bright light which lit the dummy up bright green. She then turned the house lights up and had everyone turn around and look at the balcony rail light...it was in pure white, no gel (turning the house lights on rebalanced our eyes to a true white). Beverly had gone through a series of steps that color balanced everyone's eyes to where magenta looked pretty darned white and hitting the dummy with the white light made it look very green to everyone. That's because magenta is red plus blue and white light is red plus blue plus green and our eyes were automatically color balancing the red and blue to white so the left over color was green.

 

Final color timing for a theatrical film can be quite tricky, for instance, you have to know that in a theatre there's most likely a 6K color temperature Xenon lamp supplying the light. Apparently color print film for large release is "tuned" so that the Xenon spectrum provides the truest color rendition. I've fixed a projection color problem for someone with a 35mm home theatre that uses a projector running a 3.2K halogen lamp by suggesting they put a piece of CTB gel in front of the lens, which was a quite successful gag.

[David Mullen ASC]

Yes, your eye does compensate to some degree while watching a movie with an overall warm or cold tint -- at some point, it starts to look more "normal". If you turn down the color saturation (chroma) on your TV set and watch a movie that way, after awhile, it starts to look more like a normal level of color.

 

I think the size of the image and its reference to other colors in the space can have an effect too.

 

But also the real world is "full spectrum" so to speak when lit with full spectrum sources, whereas the real world reproduced on a piece of film no longer contains as much color information -- if we print the image very blue, for example, it doesn't have much red information hidden in it compared to your eyes looking at something in a full spectrum but blue-ish light source.

[Steve London]

Illuminating responses, gentlemen, thank you.

 

Mr. Smith's got me remembering some optical (visual, if you prefer,) illusions that I had seen that used shadows and proximity to other colors to fool the eye/brain about whether two non-adjacent colored squares were the same. So far I can't find them back, but look at this fascinating effect. The marked squares are the same shade of gray!

 

http://web.mit.edu/persci/people/adelson/c...w_illusion.html

 

[Alfeo Dixon]

The marked squares are the same shade of gray!

Dagnabit... I hate perception sometimes, but then I would hate my job... I LOVE perception

[DJ Joofa]

The eye integrates the entire visual field to auto white balance. The photo isn't the entire visual field, it's being illuminated by the color temperature of the overall lighting.

 

True, and an important factor in the retinex theory of vision. That would also imply that a theater projection will be more forgiving as far as minor color off-balance is concerned as opposed to living area TV display, since most of the light/visual information for color constancy is provided by the light from film source (discounting the flare reflected from walls), where as it is difficult to control TV area extraneous lighting.

Edited December 30, 2008 by DJ Joofa