mixed color temperature and color temp measuring

[Camila Freitas]

Do daylight stocks render more natural colors in mixed lighting?

 

 

How exactly should I use a color temperature meter when I want to measure the color temp of one specific feature?

[Michael Collier]

Does it seem like we are all of a sudden inundated with color meter questions?

 

My personal feeling on them is they are for scientists. Color should be a very easy thing. Yes your eyes balance what you see, but if you work at it, it should be a very simple thing to determine 5600 from 3200 and several places in between (assuming there ever is a question on the sources tempature, which technically you should know the temp of each source before lighting)

 

Typically you have only a few types of light in the room: Daylight ambient, which is pretty much 5600K regardless (unless of course its sun up or sundown) then theres daylight direct, which is obviously much warmer. You don't technically need to know about that, however because you probably want the sun to have a natural look, so if your ballanced or timed to around 5600, the direct sunlight will read slightly golden and be very natural.

 

Then you have your typical tungsten. There is a little more variation in those, especially if you are doing doc work or ENG. Most standard household bulbs are slightly yellow (around 2800-3000) usually this is fine if its just a practicle in question, since a little warm light won't throw things off too far. If your concerened then switch it to a photobulb for 3200. Another kind of light to watch out for is the dimmed tungsten. These are everywhere, it seems. Some hotels in an effort to be more 'upscale' seem to place a lot more lightbulbs and then dim them for a golden effect. Usually the result is orange on film.

 

Then there are your standard floros, which come in cool white, delux cool white, tungsten and 'full spectrum' (which are very common in alaska) For these bulbs it is best to refer to charts availible almost anywhere to get the greenspike out. Most bulbs vary in terms of greenspike. If your unsure try and go a little over what you expect. It is better to have the image slightly magenta than slightly green (all this is quite achedemic if you have no lights of your own, since the white balance or timing will take care of most problems)

 

Thats just to get white light. Now if you want your light to be a different color or have a gradient effect, then its all a matter of feel. A color tempature meter is quite worthless, considering all the forces at work (film stock, timing, printing and development options, etc) You should keep in mind color as units over or under white. Imagine if you placed 1/2 CTB on a tungsten light. You are now 1/2 stop under white (under being towards blue), unless you were on a daylight stock, then you would be 1/2 over.

 

The reason this make things easier is for a specific stock you can remember what 1/2 and 1 stop under and over looks like in color rendition, and use that to interpolate just about any variation in between with some kind of accuracy. tempature in Kelvin is way to accedemic (at least for me) to find any use on set with.

 

In the end you really can just look at a scene and know how its going to read if you have a few points of referance (and some experience in the format).

[David Mullen ASC]

While I don't have a color temp meter, my gaffer has occasion to use one, so I'd hardly call them useless, just overrated perhaps.

 

Basically you can measure two different things on a color temp meter, the color temperature of a light, and how much green versus magenta correction it needs to be "white".

 

But the truth is that you can figure out most of your color problems with your eyeballs and an understanding of what the type of lights you are encountering, if they have unusual color spectrums, or even if they only have a partial spectrum with a particular bias. A digital still camera can also be useful in checking how a mixed color temp situation will render, although it is not an exact match to film's response.

 

You know an HMI will be daylight-balanced even though a particular one can be cooler or warmer compared to another -- but usually you can see when they are off, and generally what's important is balancing them to each other if they are supposed to match.

 

Which brings up another point, which is that mixed color temperature is not necessarily something you want to avoid or "fix" since it can be natural-looking or an interesting visual effect.

 

Where a color temp meter comes in handy is when checking a batch of HMI's during the rental check-out, for example, to see if they match well. Or when trying to match your own lights to something odd on location, like a discharge lamp or overhead flos, etc. For example, if you are filming in a warehouse lit with mercury vapor lamps and you want to know what gel to put on the windows so that the daylight background doesn't go pink in comparison once you time out the green of the mercury vapors.

 

As to whether to use daylight-balanced stock in mixed situations, some people say that they handle the mixed colors better (less exaggerated differences) although I would tend to pick the color-balance of the stock based on the dominent color temp of the scene -- if it is Cool White flos, then daylight stock may be better, but if the dominent light is coming from Warm White flos, I'd probably go with tungsten stock. Only the Fuji Reala 500D stock is particularly designed to semi-adjust for the green spike in fluorescents, but then you have to deal with the fact that it is a somewhat soft & grainy stock.