The Art of the Mired Measurement

Jay Holben

In a recent lecture on lighting, I had a student ask me about mireds and I saw the blank and quizzical faces of his fellow students at the mere mention of the word. It occurred to me that the beauty of mireds might be somewhat of a lost art, but they’re incredibly useful to understand, and a great followup to last month’s column on maintaining color temperature on-the-fly.

SOME BASIC MATH
How do we know what strength of color temperature orange or color temperature blue to utilize? It’s hard to explain that Full CTO will convert 5600K to 3200K (a difference of 2400K), but not convert 7000K to 4600K (a difference of 2400K). Unfortunately, it just isn’t that simple.

To get a better understanding of how much a particular filter will affect the color temperature, you need to get into mireds. Now this requires breaking out a little math, but I promise we won’t go beyond basic arithmetic.

A mired is a microreciprocal degree. It’s derived by dividing 1,000,000 by the Kelvin temperature. So 5600K has a mired of 180 (1,000,000/5600 = 179.57).

If you look at the manufacturer swatchbooks for lighting filters, you’ll see that each color correction filter has a mired shift factor. Lee’s Full CTO has a mired shift of +159 and 180 + 159 equals 339. Then 1,000,000/339 = 2,948, which rounds to 3,000K. So Lee’s Full CTO actually corrects 5600K to 3000K (not the expected 3200K), which is a difference of 2600K.

Gel swatch books If, as in the example above, you actually had 7000K instead of 5600K, 7000 has a mired value of 143 and 143 + 159 equals 302. Then 302 mired equals 3300K. So Lee’s Full CTO on a daylight source of 7000K would give you a color temperature of 3300K, which is a difference of 3700K! The same Full CTO filter alters color temp 2600K in one instance and 3700K in another.

A STEP FURTHER
This system of mathematical calculation is important to understand so that you can compensate properly. It can also be used inversely. If you have a color temperature meter and you read the daylight coming through a window at 9000K, you can determine that it has a mired value of 111. You want to correct that to 3200K, which has a mired value of 313. Subtract 111 from 313 and you need to have a mired shift of 202. Looking through the Lee swatchbooks, you can see that a Full CTO (159) plus 1/4 CTO (63) will get you to +222 mired 9000K to 3000K, which is close enough.

Anything less than 100K isn’t really discernible by the human eye and it really takes a shift of about 500K or more to make a noticeable difference. Anything within 500K should be totally acceptable for most applications.

Where things get a little surprising is in understanding that not all filters are created equal. A Lee Full CTO has a mired shift of +159, but Rosco’s Full CTO has a mired shift of +167 (slightly warmer than Lee) and GAM is +146 (slightly cooler than Lee). Each manufacturer has a different compensation depending on what they determine is standard real-world “daylight” and real-world “tungsten.”

Plus green and minus green are used for different applications. Non-incandescent lighting sources, such as fluorescent fixtures and HMIs, often have a tinge (or more) of green to them. Green is a challenging color because it typically is unflattering on skin tones. Generally you want to get the green out, but if you do so later in color correction, this adds magenta to the image and that can make flesh tones appear overly pinkish.

To compensate for this additional green, you use minus green filters, which are various strengths of magenta color. This blocks the green light while letting through the rest of the colors.

Inversely, if you find yourself in a situation where you can’t correct the green— perhaps you’re shooting in a huge industrial warehouse with thousands of fluorescent ceiling fixtures—then you can add green to your own lighting to even the playing field, which will make color correction go smoother later. That’s where plus green comes in.

Unfortunately, there’s no microreciprocal degrees for green/magenta shift. I’ve always found that you just need to test to find the right combination (which, in my opinion, is usually substantially less than a color meter suggests).

Jay Holben is the technical editor of Digital Video and a contributor to Government Video. He is also the author of the book “A Shot in the Dark: A Crative DIY Guide to Digital Video Lighting on (Almost) No Budget.”