Color temperatures

[Rich Hibner]

Is it possible to mix different bulbs and have good results? For instance. You wanted to light a really big room around 900sq feet with cathedral ceilings, using lights around 5000k And wanted to light a few characters with regular 3200k bulbs, to give the effect of being lit by household lamps. What's going to happen when you white balance? Is something going to get thrown off?

 

Sorry about any misplaced lingo. If I'm using wrong terms here, I'll try to explain better if you ask.

 

thanks

Edited May 17, 2008 by Rich Hibner

[Michael Nash]

You can generally only pick one color balance at a time with film or a video camera; daylight, tungsten, or something in between. Light sources that match your color balance will appear white, and light sources of a different color temperature will appear different. If you balance in between, the daylight sources will appear slightly blue and the tungsten sources will appear slightly warm, but less so than if you had balanced completely for one or the other.

 

You'll find that straight mix of daylight and tungsten will still appear a little too exaggerated on camera to appear naturalistic, so you usually try to "cool off" the the tungsten light with 1/2 or 3/4 CTB to make the temperature difference less extreme. But it's a matter of degree and taste.

[Walter Graff]

Rich you can try to mix both and white blance and see what you get. TO white blance make sure your white card sees a bit of both.

 

Are you saying you want 5k light in the ceiling area and 3.2k light for the talent? If so and you want the color contrast, white balance for the 3.2k and let the 5k go bluish which it will. Its actually a nice color contrast for teh senario you describe.

[David Mullen ASC]

Just repeating what has already been said... Mixing color temps is a nice effect and happens in real life, you just want to decide which source is the "white" one in the frame, if any, and you want to be mindful that film and video see these differences more clearly than your eye does, for various reasons. So unless you want an exaggerated difference between a tungsten practical and natural daylight, you may want to get the two closer together, by half or quarter.

 

White balancing doesn't get rid of the differences between lights of different color temps, it's an overall shift in some direction or another. For example, you can white balance to get rid of the green of some fluorescents, but the result will be that non-greenish lights in the scene will shift to the magenta (pink).

[Rich Hibner]

Rich you can try to mix both and white blance and see what you get. TO white blance make sure your white card sees a bit of both.

 

Are you saying you want 5k light in the ceiling area and 3.2k light for the talent? If so and you want the color contrast, white balance for the 3.2k and let the 5k go bluish which it will. Its actually a nice color contrast for teh senario you describe.

 

 

How do you prevent the 5k from going bluish? I don't mind it going blue, but I don't want to go too blue. Gel it with CTO?

But yes, I want the room to be really bright, and the talent to appear as if they are being hit with lamps around the room. I'm thinking 3.2k for that.

[Laurent Andrieux]

Problem with "cooling off" the tungsten lamp is that a half CTB just powers them down a full stop. I like better warming off the daylight sources. A half CTO only cuts down a 1/3 of a stop.

 

If you can gel the 5000K sources, it's the best way, to me...

 

Then, white balance in the "middle"...

[Walter Graff]

How do you prevent the 5k from going bluish? I don't mind it going blue, but I don't want to go too blue. Gel it with CTO?

But yes, I want the room to be really bright, and the talent to appear as if they are being hit with lamps around the room. I'm thinking 3.2k for that.

 

Then you'll need to color corect the 5k down to 3.2k if you want the room bright and the desk lamps ot be your main illumination while your ceiling is bright and white too.

[David Mullen ASC]

If you want full correction to 3200K, then the HMI's will need to be gelled with Full CTO, but so will any daylight windows in the room. If you want halfway correction, then 1/2 CTO on HMI's and windows. Since often tungsten practicals have tungsten light bulbs in them, they are harder to gel. There are blue-painted tungsten photofloods that get you a 4800K color though. And daylight compact flos, though they have some mild amount of green in them.

 

As for white balancing, it just depends. You could, for example, leave the HMI's and windows uncorrected, white balance to a white card lit by an HMI with 1/2 CTO on it (or a tungsten with 1/2 CTB) and get a halfway correction, so the tungsten goes a little orange and the daylight goes a little blue.

 

By the way, the "C" filter on many Sony cameras does the same thing, halfway correct daylight for tungsten, basically similar to an 81EF filter.

[Rich Hibner]

Would it be easier to use 3.2k to illuminate the room and hit the talent with something else?

 

Edit: Not easier, but more sense to. A big room like that I figured white light would be nice to fill the room with. And 3200k would add a nice effect of natural lights hitting them. It'll be dark outside so no outside lights coming in.

 

Maybe 5000k wouldn't be the right color temp to fill the room. Possibly String up China balls and use a warmer color temp bulbs in the lamps around them.

Edited May 17, 2008 by Rich Hibner

[David Mullen ASC]

If you want the coolness of HMI's, then they are more energy efficient for lighting large spaces at night. But if you don't need that much extra light level, then smaller tungstens would be better for a warm look. Chinese Lanterns would work, or tungsten balloons, or bounces, or spacelights, etc.

[Rich Hibner]

nevermind...

Edited May 19, 2008 by Rich Hibner

[Walter Graff]

Yes Rich, if you white balance to 5000k everything at 32.k looks warmer and more yellowish/orange.

[K Borowski]

White balancing doesn't get rid of the differences between lights of different color temps, it's an overall shift in some direction or another. For example, you can white balance to get rid of the green of some fluorescents, but the result will be that non-greenish lights in the scene will shift to the magenta (pink).

 

I've actually seen localized digital color correction done to a still photograph once, and, to me, the result just didn't look natural. I like to see blues in the shadows warmth in the highlights, with very few exceptions, as this looks "real" to me.

 

The original Star Trek series made pretty extensive use of colored gels over its background lights. No, the colors in the backgrounds aren't "CORRECT" but they are interesting.

 

Filmmaking gives you far more latitude in lighting or gelling in a matter that is "incorrect" for the purposes of furthering the story than does still photography. There are films that have shot tungsten stock outdoors unfiltered for a very blue cool look. "Minority Report" is intentionally improperly filtered in addition to being bleach-bypassed. So too, they filtered "Saving Private Ryan" (or maybe it was "Band of Brothers"? I forget now) so that only the blood had the correct color; they had to use an off-color blood in real life so that it would photograph the proper blood red despite the filtration on the camera lenses.

[Rich Hibner]

Hey, how did you know what I asked? Weird...because I over read what David wrote and it answered my own question. But thanks for confirming it. You can put some CTB on the 32k's to get rid of the yellowish orange looks right? That's my theory. Or some kind of gel to decrease the orange/yellow temps?

 

Off topic, but is it video that has trouble with problems like this or is it the same with film?

[K Borowski]

Hey, how did you know what I asked? Weird...because I over read what David wrote and it answered my own question. But thanks for confirming it. You can put some CTB on the 32k's to get rid of the yellowish orange looks right? That's my theory. Or some kind of gel to decrease the orange/yellow temps?

 

Off topic, but is it video that has trouble with problems like this or is it the same with film?

 

Actually, my experience is that digital and film handle mismatched colors very differently from one another. I'm not going to say that I prefer one over the other, but they are very different, so if you would say use a certain amount of gels over a light for film, in order to get the same result with HD you'll find that you'll often need a different amount of gels to get the same effect, which is a little bit confusing since, at the same ASA, the two should handle unbalanced color temperatures the same way. . .

 

So they both have "problems" (again something that isn't technically right isn't always problematic), but it will look different with one medium than the other.

[Walter Graff]

which is a little bit confusing since, at the same ASA, the two should handle unbalanced color temperatures the same way. . .

 

How film and video 'see' color compared to each other has nothing to do with ASA. It's about the process. Film uses a different process to make a picture than does video. Different mediums, not the same, nor do they see the same way. In fact film can see colors video can not easily see. And video is sensitive to color film is not as sensitive to. But ASA is not a term that applies in this case and in fact technically should be EI for motion picture film. Video dos not use such a term of EI nor ASA, but the manufactures gave in and let folks now use it although technically not based on a measurable standard so incorrect terminology for video.

[K Borowski]

But ASA is not a term that applies in this case and in fact technically should be EI for motion picture film. Video dos not use such a term of EI nor ASA, but the manufactures gave in and let folks now use it although technically not based on a measurable standard so incorrect terminology for video.

 

I'm not going to comment on this nitpicking the use of ASA versus EI, nor am I going to comment on whether or not video really has a speed; that's irrelevant and what you're saying would also apply to DSLRs which have rated adjustable speed ranges too.

 

All I am saying is that, at the proper rated color temperature, with *correct exposure* for both a film stock and a sensor, a shot of a grey card under balanced lighting with another grey card under another color temperature light in the background, should handle essentially the same way with film as with digital. I'm not talking about ultraviolet sources photographing as more blue with film or infrared sources photographing as more ruddy with digital, I'm saying there's a drastic difference between the way nonbalanced color photographs with digital as opposed to film, with digital looking more noticeable and drastic if something isn't "correctly" color-balanced.

[Walter Graff]

" I'm saying there's a drastic difference between the way nonbalanced color photographs with digital as opposed to film, with digital looking more noticeable and drastic if something isn't "correctly" color-balanced."

 

And that is because it is not as simple as a digital sensor and a film plane should see light the same way if they are claibrated for sensativity the same. Video is based on a subtractive process, different from film. They are different mediums that handle colors very diferntly by their nature. So to say that both should see the same if calibrated for sensativity is simply incorrect.

 

As for ASA, every now and then I stand up and shout about it because it is sloppy and wrong to use the term. And that is why many folks do not understand why digital sensors don't work well and as stable when using exposure indexes acoss the plane of senastivity. To propagate teh term incorrectly only adds to the confusion.

[K Borowski]

As for ASA, every now and then I stand up and shout about it because it is sloppy and wrong to use the term. And that is why many folks do not understand why digital sensors don't work well and as stable when using exposure indexes acoss the plane of senastivity. To propagate teh term incorrectly only adds to the confusion.

 

Ahem. . .

 

http://www.cinematography.com/forum2004/in...mp;#entry232859

 

What entitles me to all of the special attention? There are a lot of terms thrown around incorrectly. "ASA" isn't even correct when used referring to film speed. The "ISO" designation, a hybridization of DIN and ASA superceded it quite ome time ago. "Solarization" is often erroneously used to mean "Sabbatier Effect" in the darkroom.

 

Forgive my ignorance of the correct terminology. The only digital camera I own, a MiniDV, DOES use ASA/ISO/speed in full manual mode. So don't blame me, blame the manufacturers. The usage is widespread, almost ubiquitous. I don't know how digital sensors work, nor do I particularly care, but they are marketed as being adjustable to different speeds in many cameras; whether or not this is the case is due to spin by the marketing department or mislabeling to make cameras more appealing to experienced film users or some other reason unbeknownst to me.

[Rich Hibner]

Never got an answer, but wanted to verify. That if I wanted to get rid the yellowish orange look from the 3200k's after white balancing to the 5000k, I would gel the 3200k's with CTB? Correct?

[David Calson]

yes

[Michael Nash]

Never got an answer, but wanted to verify. That if I wanted to get rid the yellowish orange look from the 3200k's after white balancing to the 5000k, I would gel the 3200k's with CTB? Correct?

 

Yes. Full CTB converts 3200K light to 5600K, and also absorbs about 1-2/3 stops. Full CTO converts 5600K light to 3200, and absorbs about 2/3 stop of light.

 

And to avoid needless nitpicking, these are round numbers. There are always variances in the actual color temperature of light sources, and the precise correction of different gels. The main idea is that CTB converts tungsten to daylight, and CTO converts daylight to tungsten.

[Daniel Porto]

Never got an answer, but wanted to verify. That if I wanted to get rid the yellowish orange look from the 3200k's after white balancing to the 5000k, I would gel the 3200k's with CTB? Correct?

 

Correct.

 

CTB moves the color temperature from 3200K towards 5000K

 

CTO moves the color temperature from 5000K towards 3200K

 

 

I even spelt color the 'right way' for all you americans. Do you love me?

[Justin Simpson]

Does anybody have any experience with shooting tungstens with daylight film? Particularly the 7205 250D. I'm shooting a warm candle lit looking scene in an old irish pub with this. Would it come out too orange?

[Chris Keth]

Does anybody have any experience with shooting tungstens with daylight film? Particularly the 7205 250D. I'm shooting a warm candle lit looking scene in an old irish pub with this. Would it come out too orange?

 

That depends on what is "too" orange to you. Personally, I think it would be too orange. You might love it.

 

I would prefer to use a 200T film and warm up my lights by about half strength.