Florescent and LED spectrum is nowhere near flat

[Brian Drysdale]

You could check out this: Kenko International

It's a Minolta meter taken over by a new manufacturer. I had an older model, which I used when shooting film when matching tungsten lights to fluorescents and using CC filters on the camera.

[Bob Grant]

Thanks Brian.
I've been thinking this through some more based on all that's been said in this thread and it does occur to me that a basic colour meter is notgoing to provide much information when it comes to discontinuous light sources such as LEDs. We're now being swamped with an increasing number of these and as mentioned, with all sorts of claims. Some of the latest offers are quite expensive and some way to independantly evaluate their claims of "high CRI" would seem useful. One instrument I've just found is the Jaz from Ocean Optics - Jaz for Light Measurement . This seems designed specifically for the task at hand and would provide more information than just CT and CRI.
The cost of this instrument is quite a bit higher than a basic colour meter but with the LED instruments that we're now looking at costing around the same amount the investment would seem justified. It'd also have a place when evaluating the various cheap LED lights.

[Richard Andrewski]

Better check with them before you buy and see if its possible to measure a whole panel full of LEDs with the base module, somehow I doubt it based on what I know about this kind of equipment. You'll probably also need this too:

Ocean Optics - Smart, Innovative, Flexible, Solvers

The 20" model would be the only one that would seem to be large enough to work. Call them and ask to talk to someone and see if its possible to use with a normal large fixture like the ones people here would be using.

Otherwise, I think the base units as shown probably only work with just a single small LED like a simple 5mm type. You'll also need a calibration source as well like a tungsten lamp. They sell those in another part of the site.

[Guy Holt]

Originally Posted by Richard Andrewski

Beating up LEDs because they aren't in the mature stage of development (like fluorescents for example which have been around for decades), is like beating up a circa 1990s Pentium because it wasn't a multi-core processor like what we have today.

My posts are not meant as an attack on LEDs. I’m not saying that LEDs cannot be a valuable tool in certain circumstances. For a specific application, say where bulk and weight when traveling is a factor, a LED fixture offers the unique advantage of being compact and lightweight (but then so are Kino BarFlys.) Under certain circumstances, this advantage to LEDs can out weigh their shortcomings in color rendering. After all the color spectrum of Phosphor White LEDs is not horrendous. For example if you were to look at the image of the model in the cyan/blue dress lit by Phosphor White LEDs in my newsletter article, you wouldn’t say it looked bad and an undiscerning client would probably be happy with the results. But, when viewed side-by-side with the same image lit by a continuous source, you see what it should have looked like. For that reason I would always recommend a Kino Barfly over a LED light panel. They aren’t much bigger than the 1x1 LED panels, put out a lot more light, can be lamped daylight or tungsten, and offer much better color rendition - especially when it comes to flesh-tones which is critical when shooting interviews.

Originally Posted by Richard Andrewski

“As I said many times … if an LED fixture claims to be high CRI and you have color rendering or bias issues, its simply not using a high CRI LED. In other words, if you test it and proclaim to the world "this HIGH CRI fixture isn't rendering colors all that well, so I guess CRI as an index doesn't work!", you may have reached the wrong conclusion. The conclusion should have been "this fixture isn't really high CRI and therefore its not rendering colors all that well--I guess the specs were wrong. Or it could be "I'm using film so I have different issues to worry about than digital users… Again film and digital have such different behaviors that you must be more careful with discontinuous lighting for film use.”

We have to be careful in this discussion not to confuse separate issues. First, CRI ratings are not a valid measure of judging the color rendering capability of LED luminaries for photographic purposes. The CRI Index was not designed for photographic purposes, but simply to provide a reference scale for general illumination. It is simply because there is no means by which to measure for photographic purposes the discontinuous output of LED lights (CCT is for continuous spectrum black body radiators only) that LED manufacturers use the CRI index.

CRI ratings should always be taken with a grain of salt. The CRI index is a measure of the ability of a light source to reproduce a minimum of 8 mid saturated colors faithfully (a different 8 colors are used in Europe.) As such CRI can be a very limited measure of the color rendering capability of a light source; and, it is possible for a LED luminary manufacturer to game the system by tuning the output of their LED to the limited color range of the CRI color scale and achieve a high CRI rating while delivering very poor results on screen.

What makes CRI an imperfect measure? It really only tells you how a light source will perform in a narrow band of mid saturated colors, and it is best with sources that lie near to the black body line and that don’t have too high or too low a color temperature. CRI wasn’t designed as a measure for white light produced in the way that an LED does it, and when CRI is applied to LEDs it can produce misleading results. For example, because the eight indices for the individual test colors are averaged together to produce the final CRI, a light source can score well even though it renders one or two colors poorly. With a large gap in wavelengths from 465-510nm Phosphor White LEDs can do a bad job of rendering cyan but still get a respectable CRI. Additionally, because the eight standard sample colors are all of fairly low saturation, the CRI tells you nothing about how a light source will perform when rendering deeply saturated colors. The widely separated peaks of the spectra from Phosphor White LEDs can perform poorly when rendering saturated colors outside those peaks, but the current CRI definition doesn’t pick up that deficiency. The problems with CRI and solid state lighting have been recognized and a new metric, CQs or Color Quality scale, is under development by the national Institute of standards and technology (nIst) to address these and other concerns.

Second, the green bias of Phosphor White LEDs, and the greater sensitivity of film (rather than digital) to it, is a separate issue from the appreciable drop-off of wavelengths over 600nm in the output of Phosphor White LEDs. Where it is these longer wavelength colors that are so critical to rendering flesh-tones accurately, their absence in White Phosphor LEDS effect the rendering of flesh-tones in Film and Digital imaging systems equally. Where it is possible to correct the green bias of White Phosphor LEDS with minus green (magenta) gels, it is not possible to correct with gels for the color deficiencies of LEDs that effect both Film and digital imaging systems equally.

If you look at a Spectral Power Distribution graph of a White Phosphor LED light, a gel pack that would match it to a continuous light source (whether Tungsten, HMI, or LEP) would have to include a violet gel to extend its’ spectral output below 425nm. It would have to include medium blue, cyan, and turquoise gels to fill in the missing wavelength from 465-510nm. Finally it would have to include pink, red, and orange gels to extend its’ spectral output beyond its’ 600nm cut-off. All of these gels would have to be quite saturated, since there is very little, if any, output of these wavelengths in White Phosphor LEDs to begin with (especially 3200K LEDs.) Imagine how much light you will get out of a LED light panel with such a gel pack (LED light panels put out barely enough to begin with, and have no output to waste to such accurate color correction.) In other words, White Phosphor LEDS are so deficit in certain parts of the color spectrum that by the time you came up with a color gel pack to match them to a continuous light source, the LED panel would put out very little light with all those gels on it.

Originally Posted by Richard Andrewski

” … the AMPAS study was mainly concerned with film use.”

Yes, the AMPAS study was mainly concerned with film use, but there are other critics of the color rendering capabilities of LEDs. One of the best independent experts tackling the ins and outs of LED technology is Mike Wood. He has written about LEDs and other interesting things in his regular column called “Out of the Wood” published in Protocol Magazine. Links to some of Mr. Wood’s articles are listed here. For more check out his web site: Articles

When White Light Isn't White - Part 1
http://www.mikewoodconsulting.com/ar...is%20White.pdf

When White Light Isn't White - Part 2
http://www.mikewoodconsulting.com/ar...20Part%202.pdf
CRI – What does it really mean? - Part 1
http://www.mikewoodconsulting.com/ar...%20CRI%201.pdf

CRI and the Color Quality scale? Part 2
http://www.mikewoodconsulting.com/ar...%202%20CQS.pdf

Mike Wood’s articles and the AMPAS study confirm in regard to LEDs what we have known for years about general illumination light sources with discontinuous spectrums.

Originally Posted by Richard Andrewski

“As far as whether a high CRI LED can or can't render colors: when a true high CRI LED shows up, and its not all that far off, then they will render colors every bit as well as the best flos do today”

There are limitations inherent in remote phosphor technology that make it unlikely that we will ever see a Phosphor White LED with as continuous a color spectrum as the best flos do today. In the remote phosphor approach to white light with LEDs, the LED designer manipulates the color spectrum of a given LED to a desired "white light" by applying phosphor layers of distinct colors that are activated by the LED’s “Pump” color (usually the 450nm blue of an InGaN LED) to extend the color spectrum by a process called a “Stokes shift.” Depending on the chemistry of the phosphors used, the color balance of light generated by remote phosphor technology can approximate daylight, or stretched closer to a tungsten color balance. The “5500K” white LEDs use semi-transparent phosphors so the blue "pump" color comes through. In contrast to 5500K LEDs, the warmer “3200K” white LEDs have to use more phosphors and so are more opaque to the pump color, and are therefore much lower in efficiency.

There are inherent limitations to the “Stokes shift” process by which a portion of the “pump” color is transformed from shorter wavelengths to longer. First, it works in only one direction – that is why LEDs don’t emit color wavelengths shorter than their pump color and why Phosphor White LEDS, compared to continuous light sources, have no output at wavelengths shorter than about 425nm (which is why violet colors don't render well under them.)

The second inherent shortcoming to this approach to generating “tungsten” light from an LED is that the Stokes shift process reduces the total lumen output, so there is an inevitable tradeoff in lumen output and broader spectrum white LEDs. For that reason LED designers have to cut the high frequency output in the high-600 nm range. The appreciable drop-off of color wavelengths longer than 625nm that results is why pinks, reds, oranges, and other long wave-length colors tend to look dull under 3200K LEDs, compared with how they look under true Tungsten light which is a continuous spectrum light source that extends all the way out on the long-wavelength end. Since, it is these same long wave length colors that make flesh-tones “vibrant,” that accounts for why flesh-tones tend to look flat and pale when illuminated by Phosphor White LEDs.

Third, the color output of remote phosphor LEDs is very inconsistent because it is effected by the imprecise binning and manufacturing tolerances of their pump color blue InGaN LED (generally optimized between 450nm and 460nm), thermal management in the fixture, the ageing of the phosphors, and even the ambient temperature. For example, a one-degree shift in the junction temperature of the blue InGaN LED (pump color) in remote phosphor LEDs, will cause a +/- 2nm shift in the dominant wavelength. If compounded by the average wavelength variation of +/- 2nm of blue InGaN LEDs, a 5nm divergence from the prescribed 455nm wavelength of the pump color will create a color inconsistency of 5 MacAdams ellispses. While not readily apparent to the eye, image capture systems (both film & digital) will easily see this variation (and we are not talking about a green bias.)

Finally, given the unavoidable energy loss in the Stokes shift, the remote phosphor approach is not the most efficient method to achieving photographic white light with LEDS. For this myriad of reasons, the remote phosphor approach to generating photographic white light with LEDs has been all but abandoned by the more serious manufacturers of LED luminaries for motion picture production – leaving only manufacturers of LEDs for the general illumination market and those manufacturers that repurpose them for motion picture lighting applications.

No matter how much effort is spent on optimizing Phosphor White LEDS to higher light output and higher operation temperatures for the general illumination market, without a mass market for continuous spectrum remote phosphor LEDs (as opposed to high CRI – they are very different) tuned to our specific needs, there is not much hope of the prices coming down significantly. In other words, because our requirements do not coincide with those of the mass general illumination market, LED manufacturers - outside of possibly Osram - are not developing Phosphor White LEDs specifically for motion picture lighting – it is simply too small a market in the larger scheme of things. As long as that is the case, there will be higher CRI LEDs developed but we won’t see a meaningful attempt to address the color rendering deficiencies of Phosphor White LEDs for photographic purposes.

I am not saying that you can’t use LEDs successfully. As long as you understand the limitations of the tools you have at your disposal you can do good work. But, in the case of LEDs it helps to have a thorough understanding of the technology to avoid bad situations like those discussed above. For this reason, I have put together an overview of the technology for ScreenLight & Grip's E-Newsletter.

- Guy Holt, Gaffer, ScreenLight & Grip, Lighting Rental & Sales in
Boston

[Bob Grant]

Originally Posted by Richard Andrewski

Better check with them before you buy and see if its possible to measure a whole panel full of LEDs with the base module, somehow I doubt it based on what I know about this kind of equipment. You'll probably also need this too:

Ocean Optics - Smart, Innovative, Flexible, Solvers

The 20" model would be the only one that would seem to be large enough to work. Call them and ask to talk to someone and see if its possible to use with a normal large fixture like the ones people here would be using.

Otherwise, I think the base units as shown probably only work with just a single small LED like a simple 5mm type. You'll also need a calibration source as well like a tungsten lamp. They sell those in another part of the site.

I would have thought that placing the sensor at a distance from the light source would give a good reading of the combined spectrum even with sources using RGB LEDs. The measurement of total power output is of quite secondary interest. Also of some concern is the cost of that 20" sphere!

[Chris Ficek]

I couldn't agree more with Guy's excellent response to this matter. I have spent significant time and energy studying the colour quality issues of LED fixtures for the film and broadcast market and I can see from Guy's post that we have read many of the same studies and tech papers. He's right... a perfect LED does not exist, and it probably never will for the exact reasons listed. Our needs are too small for the big players like Lumileds or Osram and the such to really drop the quest for ultimate household lighting to make a good looking video light. The market for production lighting is small, probably less that $20m annually worldwide where for comparisons Phillips sold over $100m in LEDS a month in 2010, and the big LED revolution had not even started last year. You and your camera are small pickins to distract any major research effort.

Progress in LED production lighting will come from the small manufacturers that are already entrenched in our industry. They will innovate and find ways to make the LED work for us because they have something to lose if they do not evolve with the technology of LED which will surly dominate the lighting world in the immediate future.

I am in total agreement with Guy's post but I would like to but I would like to take note of his use of the term "remote phosphor". Although the use of the term in the post is 100% correct it can be misleading to those without such a well versed background on LED lighting. All white LEDs that use a phosphor are technically remote phosphor devices however the use of "remote phosphor" is often used to describe LED fixtures that use the technique of placing the colour changing phosphor outside of the actual LED itself.

This technique also called "cold phosphor" is used by Osram/Mole in their MoleLED and PRG in their new mini single LED production light. Some reviewers have given this technique very high marks for colour rendering and I do personally agree that the above mentioned fixtures are probably among the best LED colour renderers I have ever seen.

So for those of you that only read the product literature understand the difference as it is a significant one.

Cheers
(BTW I promise to finish up over at the Colouroflight.com soon, I'm the world's least prolific blogger I guess)

[Michael Panfeld]

Originally Posted by Brian Drysdale

You really need to test with the meduim you're using. Video cameras have traditionally been pretty blind towards the green in standard fluorescents, whilst with film the green does come out pretty strongly. You need to test with the camera you're using to see how it's responding with the greens, some of the new digital cameras may pick up green because they not limited to rec 709.

Very true. And its not limited to the camera sensor. Lenses do not equally transmit all wavelengths of light. If you take a lens and put it on test equipment such as a Richter autocollimator and push different colored light through it, you will see differences in the collimation points and differences in the ability to resolve light on a focused plane. True, they are small, but they are there. Same thing with a Lens Test Projector. These differences exist between brands and also within a series for one manufacturer. For old lenses, there are differences between two copies of an identical lens.

The bottom line is that each situation is unique. There is a reason that DPs shoot test footage as part of pre-production. Each step may introduce a variation, from the light source, to the actors or the scenery's reflection, to the lens, the sensor or the film type, through the chemical processing or digital treatment.

[Michael Panfeld]

Bob: check out the thread Color Meter Weirdness. There is a lot of technical info. I started that thread. I have the Sekonic 500C. What I can tell you is that the CT diffs exist between film and digital modes for discontinous light sources. For the sun, tungsten, and HMIs, they are the same (more or less, if you use a meter you will note that even a fraction of an inch in movement will change the reading).

The digital mode is basically a computed calculation that attemps to approximate the Correlated Color Temp for LEDs/flouros. Is Sekonic's algorithm correct? What are their assumptions? Who knows. But its better than nothing, even though it may not be accurate. At a minimum, it can be used to compare different LEDs or compare two examples of the same LED to see if they are consistent.

[Michael Panfeld]

I can't believe that nobody, especially Guy Holt, has yet to mention the newest and potentially the best solid state lighting technology: LEP or Light Emitting Plasma. This seems to be an answer for a lot of the LED drawbacks, while providing a more efficient use of power than a tungsten or even an HMI lamp.

Its still a bit pricey and has limited distribution, fixture types, and wattage options to date. But it clearly has a lot to offer, especially in regards to the spikey color spectrum of LEDs. Hopefully, there will be a price drop after the Early Adopters clear out the R&D investments and manufacturing ramps up.

[Chris Ficek]

I have seen LEP in action in entertainment fixtures and personally (electric considerations aside) I'm not sure if it will find much traction.