Accuracy of various lights

Phil Rhodes · July 29, 2016

So I've been messing about with lights and bits of gel and thought I'd ask what everyone thought about some results I'm getting.

I'm looking into the use of technical CTO/CTB filters and the accuracy of things like "daylight". First I talk about how HMI is viewed as being "daylight" despite the fact that one I own reads at about 6500K and should really have a bit of 1/8 CTO in front of it at all times.

I shot a macbeth chart in both clear, midday daylight and HMI and of course they don't match precisely. Then, I put some full CTB on tungsten and shot the chart again and of course they don't match again. The tungsten actually looks like it's behind about one-eighth plus green, which is something I've seen before. Tungsten can be greenish, for some reason.

Now in my experience this is pretty normal - things get lumped into categories like "tungsten" and "daylight" and of course they often don't really match, but still.

What has everyone else found?

P

Mark Kenfield · July 29, 2016

My experience is that, for a consistent 'daylight' from assorted lights (tungsten with ctb, tungsten with dichroic filters, HMIs, Kinos, LEDs) you have to test and design filter packages that bring everything in line.

Stuart Brereton · July 29, 2016

Daylight changes a lot. In clear conditions, it's a mix of Skylight which is around 10,000k and Sunlight, nominally around 5000K, but which varies according to time of day & year, location and how much atmospheric pollution it's coming through. In overcast conditions, that same mix is coming through hundreds of feet of dirty water vapor, which is doing who knows what...

Then you have the problem that lamp manufacturers can't seem to agree on a CT for daylight, so you have lamps ranging from 5000k to 6500k, which all look a bit like daylight under certain circumstances.

Lastly, not all CTB gels are created equal. If you look at LEE CTB and Rosco side by side, you'll see that the Rosco gel leans toward the magenta, whereas the LEE is more neutral, although you could argue that the Rosco gel is correct and the LEE is more green, I guess.

Working here in southern California, I'm pretty much guaranteed that 'Daylight' means sunlight, so I tend to use HMIs with 1/4 CTS as a main source, and then use daylight kinos as a slightly cooler fill. I try not mix lamps, so if I'm ever using Tungsten with CTB as daylight then that's all I use.

Satsuki Murashige · July 29, 2016

Not to mention, not even all tungsten lamps are perfectly consistent. Between the same units, the color can vary due to age or type of globe. Kino tubes of the same make and type can be visibly off from each other due to age as well.

Probably would be best to line up all your lights against a long white wall in a dark room and photograph the results to see what the differences are, then adjust with gels from there. As long as the lights all match each other, that's a good start.