A Lesson in Colour from Dedo Weigert

In an earlier post regarding the new Dedolight LedRama, I got slightly confused and stated that future versions may have a green/magenta control dial in future (removed from post). 

Dedo Weigert from Dedolight read this and took the time to explain why they definitely won't have this feature and how they have developed the tech in their new series of LED lighting. It was such a fascinating explanation, I had to share it! This is why I put so much faith in Dedolight products!

Here is Dedo's email:

"We believe in something that is called the Planck Curve, and that is where the sun lives, where halogen lives, and where all good lights live.

Household fluorescents deviate from the Planck Curve quite seriously. Kino Flo is very close to the Planck Curve, at least with the T12 daylight lamp.

Regarding color -

You will find a lot about CRI, which is a system from 1931.

Now we use expanded CRI, measuring 15 colors with a spectrometer. Again, that is only one step closer to the reality of the digital camera.

Alan Roberts has developed a system called TLCI (Television Lighting Consistency Index), which is a huge step closer to the real world of digital cameras. In my mind, however, it is based on studio cameras which usually work with 3CCD sensors.

Many people roaming around and small teams will never use such cameras, they all work with CMOS cameras. The devil has it that each CMOS camera reacts differently to LED light. They all see daylight pretty much the same, they see halogen light the same, to some extent also HMI and fluorescent, but with LED they show vast divergences and different reactions.

This is why we shoot endless tests, taking people’s faces, which in my opinion is close to the reality of our tasks, where we light one side with a reference light, be it halogen at 3000K, 3200K or 3400K, and the other side with our LED lights and with other people’s LED lights.

Then we find that rarely ever the two sides of the face show the same skin color.

We are now very happy that we have come to a stage, where with many different CMOS cameras from Alexa to RED, to Nikon, to Canon, to Sony, to Black Magic we have been finding total congruence on both sides of the face.

That is what he have been striving for, and for a long time I thought we will never reach it, because even cameras from the same manufacturer with the same model number may show different reactions to LED light.

To give you an example -

The Canon 5D MarkII needed a correction of 1/8 minus green, and since you are working with LEE, you are well aware what that means. The Canon 5D MarkIII needed a correction of 1/4 minus green, whilst the Sony cameras – at least some of them – needed a plus green.

As long as you have LED light sources that cause these problems, you may dream about a correction in the green / magenta scale.

With our lights we are now absolutely on the Planck Curve and the deviation from the Planck Curve we keep at absolute minimum tolerances. These deviations are called Delta UV, and the maximum tolerance that we accept is 0.0054. Usually, however, we are much closer, and beyond that we now start to identify each and every individual light by this deviation in Delta UV measurements. So even if there are minimal differences between the lights, we can identify them, and if somebody wants to reorder lights, we can supply lights that give exactly the same color.

I am very proud that after so many years of intensive development we can now say that our LEDs, over a long time in tests we are doing for 10000 hours or more, will show a little shift in color temperature, where we find that the daylight color temperature after many hours of continuous use in maximum heat will go up in color temperature a little bit, whilst the tungsten LEDs under the same extreme conditions will go down a little bit in color temperature.

The wonderful thing, however, is that they still stay in the same 'relationship‘ to the Planck Curve.

Why do we need green, why do we need magenta?

Both of those colors are absolute killers for skin tones. We want to avoid them and run away from those as fast as we can. I am very proud that we are right on the Planck Curve and that we can prove this in the endless amounts of skin tone tests that we conduct even with different skin characters.

We make these tests with Scandinavian skin tones, Mexican, Ethiopian and Angolan skin tones. Everything from the palest to the darkest, because every experienced cameraman knows that different skin structures and skin tones may react differently.

If you ever shot in Africa and had a very dark person in front of you, you may have found that on some of them the back light turns blue, whilst on others with similar skin tone color the back light may not turn blue at all.

Therefore we are very proud to have found congruence and consistency in skin tone results. That, in our profession is the most tricky task, because it is very easy these days to have similar light sources all over for key light and fill light, and then any deviations can be fixed in post production. But when you mix light sources with traditional light sources, I am very proud to say that we are finding very good matches, and we are still working on decreasing the tolerances, getting away from the green, getting away from magenta, staying consistent and keeping the same color for longevity without leaving the Planck Curve. 

In some of these long-term tests we even found that after 5000 hours or even 10000 hours the Delta UV value has even improved (has gotten smaller).

So why would I ever want to include something that gives me a beastly green or magenta, when, for many years now I have lived in the dreams of getting away from these killer colors.

That is my personal opinion, but I am quite sure that I am not the only one who has this opinion, striving for clean skin color, and even in the most difficult tasks make lights match with other traditional clean color light sources."

Thank you Dedo for once again taking the time to provide such an educational response!!!