Cinematography Mailing List - CML

LED Fixtures/Color Rendering

I had an odd enquiry today about the use of LEDs as a compact light source in a film application. The device was required to have a colour temperature similar to that of tungsten, but while I know that there are "warm white" LEDs available, I'm not sure how they would show up on film, since the light output is a combination of a blue LED and the phosphors that it stimulates. I guess that this means that the spectral output will be rather peaky with no guarantee that there's not going to be some ghastly tint caused by spectral incompatibility with standard film emulsion.

Has anyone had any experience with white LEDs (even the cold white ones) on a film application?

Clive Mitchell


The LED frontier is growing every month and will rapidly become an important option for many lighting applications. This month one of the main manufacturers of LEDs (Lumileds) lumileds will begin shipping their new "white" LED with a CCT (Correlated Color Temperature) of 3200K and a CRI (Color Rendering Index) of 85+.

That makes it a good match for nominal incandescent and with CRI numbers (but not necessarily color temperature) that are about on par with the best fluorescent and HMI sources.

Here's a link to their press release on the topic :

This new product (which may be sold under the name Luxeon) was only recently announced and presented at the recent Lightfair International trade show and conference in New York. The lumen maintenance is said to be 70% of original design output levels through 50,000 hours of use. That's a big improvement. I have not seen any spectral analysis of the LEDs outputs but it's said to be much better in the red range than previously available. I can't wait to see some of the light fixtures that our industry to build around this new source. I'm looking forward to seeing Walter Graff's work on his "sun gun".

One concern that has not yet been well addressed is the efficiency of these high CRI LEDs. While the lumen-per-watt output of the red and yellow LEDs is quite high, the output for the high CRI "white" LEDs is not much better than incandescent. If I read their posted information correctly, the output of the new LEDs is about 22 lumens per watt. That's only marginally better than an incandescent source, but the technology keeps getting better every month. If you're interested in new light sources, this is one watch.

Bruce Aleksander
Houston, Texas

Great post Bruce. You've obviously been watching the LED market.

>If I read their posted information correctly, the output of the new LEDs is about 22 >lumens per watt.

The secret is in proper lenses and reflectors. An LED by itself is well, just an LED just as a 1000 watt lamp isn't much fun without a fresnel to shape the light it produces. But add the right reflector and/or lens to the design of an LED and you've got an efficient piece of light. Take an LED with a flux of about 25 and add an efficient form of lens like the Luxeon Emitter does and you get a flux per LED of over 100 lumens. That's better than an incandescent light. Gang up a few of these puppies and you have a light source that needs little power, creates zero heat, has instant on capability, and can be used in methods never imagined.

My sun gun was my first experiment using nothing more than Luxeon LEDS mounted in the end cap of a 4 inch piece of PVC with a variety of snap in diffusers. I am now working with a number of different Luxeon LED configurations. I hope to create some sort of usable studio fixture from LEDS shortly. I'm building various designs right now and watching as efficiency continues to improve.

If anyone has any custom applications let me know.

Walter Graff

WalterNY writes :

>light source that needs little power, creates zero heat, has instant on capability, and >can be used in methods never imagined.

Creates _LESS _heat! The Luxeon stars MUST be mounted on a suitable heat sink for continuous operation at high level. The aluminium plates that they are available pre-fitted to are simply a convenient means of interfacing the LEDs to a larger heat sink. Using the Emitters directly requires that their base plates be isolated from a common heat sink by a shim of thin fibreglass laminate bonded to an aluminium substrate.

If the Luxeon stars are not adequately cooled, then their life will be shortened. Purely passive convection cooling would require a fairly chunky heat sink if a large cluster were to be built, so a small cooling fan running at low speed for noise reduction would be beneficial and make the units more reliable.

I tested some cheapo Taiwanese white LEDs in a video environment today, and was surprised at how well they took gel. Then again, video tends to be a bit more forgiving than film.

Clive Mitchell

>Creates _LESS_ heat! The Luxeon stars MUST be mounted on a suitable heat sink >for continuous operation at high level.

Systemically they create more heat. Hence the reason why they can run three times the current through these LEDS and the only reason they can get so much light out of them. My reference was in comparison to radiant heat.

Someone here last year wanted to know how to shoot ants at high speed with lots of light. The answer would have been an LED bank which would generate ZERO heat to the ant and about 500 foot candles for the camera. Try that with an incandescent fixture and you've got roasted ant.

Walter Graff

>Has anyone had any experience with white LEDs (even the cold white ones) on a >film application?

I've done extensive testing of various LEDs and film color rendition including the latest generation 3200k versions since I am working on developing lighting fixtures using LED technology. I have created a working sun gun that I am currently using in the field with great success.. I will find the tape I made of some film tests take some stills and create a web page this week. The spectral graphs of LEDS are a lot less spiky than you think although lacking in some bands of color.

Walter Graff
Amherst, MA

WalterNY writes :

> The spectral graphs of LEDS are a lot less spiky than you think although

I guess that the CRI could vary widely between different manufacturers according to their choice of phosphors. Nichia aren't just a primary manufacturer of white LEDs, they also manufacture specialist phosphors for other applications too, so I guess that theoretically they have the most expertise in this area. Then again, Agilent seem to have made the first big leap with their 1W and 5W Luxeons and their different colour temperatures.

How were the results of your experiments? Was the colour good or did it have a bias?

Clive Mitchell

Bruce Aleksander writes :

>If I read their posted information correctly, the output of the new LEDs is about 22 >lumens per watt.

Bear in mind that the output from the LEDs is much more controlled than from an incandescent lamp, so combined with the relative efficiency, the LEDs are quite interesting.

I hope the fixture manufacturers have learned from the early attempts at high output LED products, where the design was led by salesmen who encouraged the designers to force the LEDs to the point of early failure to achieve unrealistic light output levels.

Also beware of the usual marketing hype that suggests that fixtures like a moving yoke light with 12 each of 1W red, green and blue Luxeon stars can compete with an HMI fixture. (Hah!)

Clive Mitchell

The LED lighting fixtures I'm currently aware of here in LA are :

Concept Lighting's Spectlight: An RGB LED light with a very nice auto sensing power supply. The power supply is of course flicker-free and dimmable. It has presets for 2900, 3200, and 5600K. It also has presets for R,G,B,C,M,Y.

There are a few drawbacks such as minimum illumination distance (before you start seeing the individual colors), output over life for different color LEDs (they don't all decrease in output at the same rate so I'd guess you'd have to reprogram the power supply's ROM at specified intervals) and lack of punch...You could probably call Hector at Concept (Located in Sun Valley) and demo one for a few days...

Fisher's Ring Light: Someone more familiar with this light could chime in, and there's probably more information on the web. It uses "white" LEDs and usually comes with a set of pre-cut gels. I've seen it come to set a few times, but never in actual use.

Color Kinetics line of LED fixtures: The fixture I'm most familiar with is the ColorWash, a 4-pin (may vary as pin assignment is user configurable) DMX controlled RGB fixtures using 12 Luxeon (5 watt I think?) ChromaCore emitters for each color, 36 total. Quite a bit of punch- relatively speaking. The fixture is about the size of a 1K nook light and originally designed for architectural lighting. Color Kinetics make a whole line of LED based fixtures including a 3ft strip light and little "i cove" fixtures designed for accent lighting. It's a pretty cool system, if it fits your application. Multiple fixtures run off a single power supply- different sized supplies are available to accommodate different numbers of fixtures.

I've never sat down and tried to dial in different color temperatures, since the applications I've seen them used for are usually color scrolling and saturated color effects. If you're LA based and get a chance to attend a function at the Hollywood and Highland Annex Ballroom (across the courtyard from the main Ballroom on level 5, just above the Highlands) there is a permanent installation using these fixtures working above a false ceiling of bleached muslin and 4x8' milk glass. The end effect- the color of the ceiling scrolls through primary and secondary colors. The Color Kinetics line can be pre-addressed with a hand held unit called a "zappy" sp? There's also a non-console control unit that has a few pre-programmed effects of limited use. I know that ELS has these fixtures for rent and you could probably schedule an in-house demo.

There's also a Canadian company called TIR systems who make an RGB fixture called the "Destiny Colorwash" which uses 60 Luxeon 5-watt emitters. I've never seen one. Prudential Lighting Products here in LA is listed as a distributor.

For a peek into the future of solid state lighting there's an interesting article on "Quantum Dots" which seem promising in overcoming many of the disappointing features of today's "white" LEDs- specifically inefficiency and color rendering.

The link is from :

Luxeon claims that by 2020 or so LEDs will be the most efficient light source available. Lumileds, which partnered with Philips to create the Luxeon line (correct me if I'm wrong), is named in the article as being part of the project with the DOE... Hopefully the technology will pan out. The DOE also helped develop the sulfur/argon microwave light that hasn't seemed to make inroads into practical we'll see.

As always the list is a great resource I really enjoy.

Stephen Olsen
Lighting Technician

Copyright © CML. All rights reserved.