CFL Color Temperatures

[UnknownVT]

On the surface this may seem simple enough -
most CFL (Compact Fluorescent Light) have a color temperature clearly marked on their packaging


However in the June/2012 issue of Popular Photography there was this article:


It would appear according to their measurements all the CFL bulbs they had,
seem to have a higher color temperature than marked - by quite a significant margin.

I have been using CFL for lighting of my photography and I had been seduced by this:

The CIE positions D65 as the standard daylight illuminant:

[D65] is intended to represent average daylight and has a correlated colour temperature of approximately 6500 K. CIE standard illuminant D65 should be used in all colorimetric calculations requiring representative daylight, unless there are specific reasons for using a different illuminant. Variations in the relative spectral power distribution of daylight are known to occur, particularly in the ultraviolet spectral region, as a function of season, time of day, and geographic location.
—ISO 10526:1999/CIE S005/E-1998, CIE Standard Illuminants for Colorimetry

from Wikipedia on Illuminant D65

So even though the 5000K CFLs seem to look better to me -
I use 6500K GE CFLs for photo illumination -
I think for fixed Daylight balance seem a bit blue'ish
(although I am very aware of it, I don't seem to see any green cast alluded to in the PopPhoto article)

Using AWB (Auto White Balance) the photos seem just fine
and in pp (post processing) I can use white point selection to correct for any color casts -

I just figured that a bit more blue may make it easier for the camera anyway.....

Still it was interesting reading that Popular Photography article -
and that could explain why the 6500K CFLs seem blue'ish and the 5000K CFL seem so pleasing to me.

I did a whole bunch of "color balance" test photos in the thread:

GE sunshine 5000K CF

[blasterman]

The reason that CFL's tend to drift to the high side of CCT is simply due to cost and marketing. Less phosphors make them cheaper to make, and the higher the CCT the higher the visual lumen measurement.

The reason you can get away with 6500k in a CFL for still photography is digital sensors aren't that demanding in terms of color source given they do significant interpolation of RGB color data to begin with. Go back a decade or two and try shooting some 4x5 transparency film with a 6500k CFLs and you would have all kinds of problems that coulnd't be fixed with any combination of gels. Even 95CRI fluorescent tubes were considered primitive compared to typical 100 CRI electronic strobe.

Pop Photography shoulnd't throw stones - they don't seem to mention the critical aspect of *CRI*, but given the mags reputation for having bad color science I'm not surprised. At least they are no longer telling people to use high contrast, low quality grocery store print films -vs- professional ones for weddings and portraiture.

[UnknownVT]

The reason that CFL's tend to drift to the high side of CCT is simply due to cost and marketing. Less phosphors make them cheaper to make, and the higher the CCT the higher the visual lumen measurement.

Many thanks for the input -
what is the native/natural color temperature for fluorescent lighting?
from: Wikipedia entry on Fluorescent Lamp - section Color Temperature

Color temperature
Main article: Color temperature

Correlated color temperature (CCT) is a measure of the "shade" of whiteness of a light source, again by comparison with a blackbody. Typical incandescent lighting is 2700 K, which is yellowish-white. Halogen lighting is 3000 K. Fluorescent lamps are manufactured to a chosen CCT by altering the mixture of phosphors inside the tube. Warm-white fluorescents have CCT of 2700 K and are popular for residential lighting. Neutral-white fluorescents have a CCT of 3000 K or 3500 K. Cool-white fluorescents have a CCT of 4100 K and are popular for office lighting. Daylight fluorescents have a CCT of 5000 K to 6500 K, which is bluish-white.
High CCT lighting generally requires higher light levels. At dimmer illumination levels, the human eye perceives lower color temperatures as more pleasant, as related through the Kruithof curve. So, a dim 2700 K incandescent lamp appears comfortable and a bright 5000 K lamp also appears natural, but a dim 5000 K fluorescent lamp appears too pale. Daylight-type fluorescents look natural only if they are very bright

The reason you can get away with 6500k in a CFL for still photography is digital sensors aren't that demanding in terms of color source given they do significant interpolation of RGB color data to begin with. Go back a decade or two and try shooting some 4x5 transparency film with a 6500k CFLs and you would have all kinds of problems that coulnd't be fixed with any combination of gels. Even 95CRI fluorescent tubes were considered primitive compared to typical 100 CRI electronic strobe.

Don't I know it, from years of shooting slide/transparency film.

But I didn't realize electronic flash could be CRI=100?
I thought real daylight was hard to match.

Film photography had traditionally set 5600K as "Daylight"

Pop Photography shoulnd't throw stones - they don't seem to mention the critical aspect of *CRI*, but given the mags reputation for having bad color science I'm not surprised. At least they are no longer telling people to use high contrast, low quality grocery store print films -vs- professional ones for weddings and portraiture.

from same Wikipedia entry:

Color rendering index
Main article: Color rendering index
Color rendering index (CRI) is a measure of how well colors can be perceived using light from a source, relative to light from a reference source such as daylight or a blackbody of the same color temperature. By definition, an incandescent lamp has a CRI of 100. Real-life fluorescent tubes achieve CRIs of anywhere from 50 to 99. Fluorescent lamps with low CRI have phosphors that emit too little red light. Skin appears less pink, and hence "unhealthy" compared with incandescent lighting. Colored objects appear muted. For example, a low CRI 6800 K halophosphate tube (an extreme example) will make reds appear dull red or even brown. Since the eye is relatively less efficient at detecting red light, an improvement in color rendering index, with increased energy in the red part of the spectrum, may reduce the overall luminous efficacy.^[19]
Lighting arrangements use fluorescent tubes in an assortment of tints of white. Sometimes^{[weasel words]} this is because of the lack of appreciation for the difference or importance of differing tube types.^{[citation needed]} Mixing tube types within fittings can improve the color reproduction of lower quality tubes.

Phosphor composition
Some of the least pleasant light comes from tubes containing the older, halophosphate-type phosphors (chemical formula Ca₅(PO₄)₃(F, Cl):Sb³⁺, Mn²⁺). This phosphor mainly emits yellow and blue light, and relatively little green and red. In the absence of a reference, this mixture appears white to the eye, but the light has an incomplete spectrum. The CRI of such lamps is around 60.
Since the 1990s, higher quality fluorescent lamps use either a higher CRI halophosphate coating, or a triphosphor mixture, based on europium and terbium ions, that have emission bands more evenly distributed over the spectrum of visible light. High CRI halophosphate and triphosphor tubes give a more natural color reproduction to the human eye. The CRI of such lamps is typically 82–100

There's a whole panel of Fluorescent lamp spectra below that section which shows how spiky fluorescent spectrums are.

[tickled]

But I didn't realize electronic flash could be CRI=100? I thought real daylight was hard to match. Film photography had traditionally set 5600K as "Daylight"

Photography literature says xenon sources have CRI ranging from 95-100.

[blasterman]

Awhile back I had the chance to do some contract work for a company that had several thousand square feet lit with very high CRI 4-8 foot fluorescent tubes. This was used primarily for Q/C purposes in clothing made over-seas and brought in for inspection. If you're used to typical office fluorescent lighting the environment under the high CRI tubes was like stepping into Oz when the film switched from B&W to color. I've wored on inspection tables that have local high CRI lighting, but an entire plant floor converted to the expensive tubes was pretty interesting to say the least. I felt bad for workers who spend 6 hours a day working under the lights and then had to move to their typical office cube for end of the day paperwork.

I'm not really a fan of 5000k or higher lighting for home use, but if you are, and primarily use CFL you really need to treat yourself to some of the specialty 'full spectrum' daylight CFLs that use the more exotic phosphor sets. Not cheap, but they are worth the trouble and some go as high as 98 CRI.

Because of the spikes of fluorescent lighting (and HID) I agree with a lot of other people that the smoother slopes of LED sources often look a lot better when the fluorescent has better CRI (on paper). I have a few dozen 5000k XP-Gs that even with a CRI in the 70's look better than any vanilla CFL in the CCT range and blend much better in a room.

Oddly, Taco Bells in my area have been outfitting their dinning areas with ~5500k high CRI tubes. Still scratching my head at that one because it doesn't make for the most appealing eating experience.

[UnknownVT]

I'm not really a fan of 5000k or higher lighting for home use, but if you are, and primarily use CFL you really need to treat yourself to some of the specialty 'full spectrum' daylight CFLs that use the more exotic phosphor sets. Not cheap, but they are worth the trouble and some go as high as 98 CRI.

I am similar - for household lighting I use 2700K CFL that seem to be like regular incandescent -
I know there is suppose to be a difference -
but I do kind of downplay the CRI=100 of incandescent (by definition,
since the black-body standard at that color temp is the incandescent light itself - so kind of self-fulfilling.
If it is so "perfect" - why is it I can't see yellow print on white, or distinguish navy from black?)

I actually prefer 2700K CFL over CRI=100 incandescent - as there seems to be a bit more blue in the spectrum so that the normal incandescent difficulties are not as exaggerated.

I also have some GE Reveal CFLs - those seem to exaggerate the reds - eg: Cheddar cheese seem way too orange.

That is why I use 5000K as a sort of inspection light for viewing photos - a cheapo way of simulated daylight.

My kitchen under-counter light was also a 5000K CRI=90 Chroma 50 tube - I prefer to be able to see my food colors properly - but when the ballast became unreliable I replaced the fitting with a standard screw in bulb holder and now have a 60watt equiv 5000K CFL (no CRI rating - it's a cheap Utilitech from Lowes)

to be honest, I can't see much, if any, difference.

I only use the 6500K CFLs for photography - fortunately I do not have to have critical color rendition/accuracy - and most of the still objects I take photos of under those lights - seem to be OK - but then I do use AWB when shooting - fixed daylight balance would tend to make them a bit blue and 5000K would seem "better".

It may seem *ss-backward logic - but is it possible in balancing a light that's a bit blue the camera has to compensate with more warm colors like yellow/orange/red?

EDIT to ADD - found this under Color Rendering Index on Wikipedia:

Film and video high-CRI LED lighting incompatibility
Problems have been encountered attempting to use otherwise high CRI LED lighting on film and video sets. The color spectra of LED lighting primary colors does not match the expected color wavelength bandpasses of film emulsions and digital sensors. As a result, color rendition can be completely unpredictable in optical prints, transfers to digital media from film (DI's), and video camera recordings. This phenomenon with respect to motion picture film has been documented in an LED lighting evaluation series of tests produced by the Academy of Motion Picture Arts and Sciences scientific staff.^[30]

[UnknownVT]

EDIT to ADD - found this under Color Rendering Index on Wikipedia:

Just a follow up -

Using the link at Wikipedia take us to this page - Solid State Lighting Project
very interesting videos from the symposium.

It appears that solid state lighting ie: LEDs can cause lots of problems in film and video -
even when they may appear to the eye as indistinguishable from normal lighting for the film industry.

There are lots of videos - this is a "summary" of sorts:

Summary

This is a good example with flesh tones and makeup -

Makeup Case

[Canuke]

Wow, this thread got interesting in a hurry for me. I was at NAB in April, and noticed that solid state lighting had largely taken over the small and medium sized studio lighting market. From a durability and power consumption standpoint, I could see why; the nature of this market is such that dedicated designs that would be too expensive for the mass consumer market (where they are still trying to stuff LED's into the A19 form factor and sell it for $20 or less) work very well in niche markets with high prices and margins.

A common configuration were light banks that were 50-50 cool and warm white emitters for color temp variation; less common were some that used a mixture of cool white, warm white and/or amber, and red to simulate a wider section of the blackbody curve. There were also wash lights at one vendor that used lots of the RGB MC-E emitters, both cool and neutral whites, as well as an MC-E that had two cool and two warm dies.

But I didn't see any of them pitching "high CRI". I reasoned that this was due to a combination of factors: current cameras are already sampling at just three spectral points (R, G and B), and so will likely be unaffected by the cyan "ditch" or by the lack of deeper reds. They are also highly white-point flexible, and last but not least, the vast majority of footage is shot with the expectation of being heavily manipulated in post anyway, in color correction and grading (which almost completely eliminates the issues with gels etc. that were such headaches for film).

Apparently, the "spikiness" of fluorescent sources is less of an issue for modern cameras (probably because the sensitivity curves for each channel are wide enough to swamp out the spikes) than the smoother-but-overall-less-even LED spectra. If so, HCRI may yet enter that market as discerning directors get used to the new lights.

Interesting stuff! I would have thought that those guys would have been ahead of the curve on this point... but CPF's tint snobs beat 'em to it

[UnknownVT]

Apparently, the "spikiness" of fluorescent sources is less of an issue for modern cameras (probably because the sensitivity curves for each channel are wide enough to swamp out the spikes) than the smoother-but-overall-less-even LED spectra. If so, HCRI may yet enter that market as discerning directors get used to the new lights.

Many thanks for that input -
I remember you gave some useful input in the thread:
LED Stage lighting

That thought had occurred to me as well -
fluorescent spikiness is broader/less peaky that those from LEDs -
somewhat like gel'd incandescent lights are far less spiky/peaky than the equivalent LED -
so may be a bit more forgiving on photos.

First a caveat - I am not by any stretch of the imagination a critical color user - so may well miss some of the finer nuances in color rendition - ut here are some of those Macbeth panel photos:






Thanks,

[blasterman]

But I didn't see any of them pitching "high CRI"

Likely because it doesn't make a difference. The difference between lets say a neutral LED at 80 CRI and a neutral LED at 90 CRI is typically just the red/amber ratio. It makes very little difference with digital capture and is easily adjusted for.

What concerns me about the 'Solid State Lighting Project' links is there are no details about the products used other than vague references. The vast majority of solid state video lights are composed of 5mm and 10mm LEDs, and we all know the wonderfull color quality of cool white Chinese LEDs in that category. Funny seeing the charts of the RGB LED used, which shows the limited gamut of tri color LEDs.

Using pretty much any neutral white LED with a CRI of 80 or above will easily compete with metal halides. The problem is they are stuck in a 5500k universe where the only other alternative is tungsten. If you don't want anemic red response then don't use low CRI cool white LEDs.

[blasterman]

Oh yeah,,,,,it would be really cool to see a typical LED retrofit bulb thrown in here for color sake. I'd actually prefer to see a warm white Ecosmart compared, and then WB adjusted.

[fnj]

(Quoted from Wikipedia: At dimmer illumination levels, the human eye perceives lower color temperatures as more pleasant...

What a load of bunk on Wikipedia's (and Kruithof's - see Wikipedia article) part. It's subjective. Daylight white is variously gauged at 6000-9000 K or so and to me is subjectively glorious beyond description.

Driving by a house with incandescent warm 2700 K lighting in the evening looks so inviting, like a Norman Rockwell painting, but I want the lighting I live with preferably in the daylight spectrum, no matter what the illumination LEVEL. It's all subjective. Maybe it has something to do with me living at 42 north. The whole winter is cold, snowy, grimy, gloomy, and grim. Look up Seasonal Affective Disorder (SAD).

[UnknownVT]

What a load of bunk on Wikipedia's (and Kruithof's - see Wikipedia article) part. It's subjective.

Precisely it's subjective -
however the Wikipedia description of the Kruithof curve is that it is supposed to be empirical -
results through observation or experimentation - ie: it seems to work for most people -
although an individual or even a group of people may see things differently,
which is a very valid observation -
unless it is in the majority - unfortunately still does not in any way invalidate a claim of being empirical.

However I take dibs on the top bunk.....

But like you, I prefer to use these 5000K and 6500K CFL's for looking at/examining color.

[blasterman]

Daylight white is variously gauged at 6000-9000 K or so and to me is subjectively glorious beyond description.

If it's at 100 CRI, which is a 'yes' for sunlight and 'hell no' for the CFL's listed here.

The 'daylight' 6500k CFL UnknownVT posted above has about as much relation to natural sunlight as a Big Mac does to Weight Watchers. The CFL out put is nothing more than a blue spike, a smaller orange spike, and a bit of yellow-green. That's not sunlight. I've never met a living human being that's tried these in their living room or dining room and either removed them out of disgust or was threatened with a divorce.

5000-5500k high CRI is nice to work under during the day, but I'd much rather have 3000-3500k LEDs in my house in the evening.

[fnj]

If it's at 100 CRI, which is a 'yes' for sunlight and 'hell no' for the CFL's listed here.

The 'daylight' 6500k CFL UnknownVT posted above has about as much relation to natural sunlight as a Big Mac does to Weight Watchers. The CFL out put is nothing more than a blue spike, a smaller orange spike, and a bit of yellow-green. That's not sunlight. I've never met a living human being that's tried these in their living room or dining room and either removed them out of disgust or was threatened with a divorce.

What about real wide spectrum fluorescents like http://www.fullspectrumsolutions.com...8_detailed.htm or http://www.fullspectrumsolutions.com...t_411_prd1.htm ? These are 5500-5900 K, CRI 93+. Lamps like the former are very successful in office lighting. I have one of the latter on the way to compare to my Blues Buster incandescent. Will report.

5000-5500k high CRI is nice to work under during the day, but I'd much rather have 3000-3500k LEDs in my house in the evening.

The idea that the human organism "wants" different CCT at different times in the diurnal cycle is a good one to explore. It opens the interesting subject of variable CCT lighting; either duplicate sets of lights or perhaps tricks using RGB LEDs with a controller.

[blasterman]

Lamps like the former are very successful in office lighting.

Which is kind of my point. I work in all kinds of corporate offices doing consulting, and at 8pm at night the last thing I want to be reminded of is having to go to work the next day, but I'm not a fan of tungsten either. If having 'daylight' or high CCT tubes in your evening living space makes you feel better by all means use them.

No matter what the CRI is of an artifical light source it can't replicate the spectrum of the sun, especially fluorescent tubes. Plasmas are close, while high CRI metal halide (we keep forgetting metal halide still exists) to my eyes replicates the sun better than any tube. If you wanted to replicate the sun's spectrum as close as possible in your living room, and coulnd't afford plasma you could always use high CRI halides in a torche' style fixture and would get really close. New ES series Rebels (5000k - 80/85CRI) are next on the list. The smooth spectrum of LEDs is something fluorescents can't match, but high CRI and high CCT is something LED's have only just started doing well.

[UnknownVT]

I did a google for CFL & photography,
and found these threads that might be interesting:

Shooting with Compact Fluorescent Lights (CFLs)

CFL 5000k vs 6500K

these were old from DVInfo.net (Digital Video Information Network)
the posts are obviously, like here, not definitive but from enthusiasts in digital video
as stated in Post #7 - it seems there are people who like using them with digital video.

Then I stumbled on a National Geographic article on CFLs

Sunshine CFL Vs. Daylight CFL | National Geographic

Sunshine CFLs
The sunshine color temperature typically ranges from 5,000 to 5,400 K (see Reference 1). This color temperature range reproduces the effect of direct sunshine. The light produced by a CFL that's rated for sunshine is generally white with hues of blue that increase with the temperature scale. Sunshine CFLs are typically marked with the word "sunshine" on the bulb packaging and include the specific color temperature. Use sunshine CFLs in rooms where you want a slightly tinted light that is softer than a daylight bulb.

Daylight CFLs
The typical daylight color temperature is 5,500 K, but most CFL bulbs that produce this type of color temperature range from 5,500 to 6,500 K (see Reference 1). This spectrum produces a clear, bright white light. Daylight temperature CFLs reproduce the effect of direct sun with lightly tinted blue sky on an average day. This makes daylight bulbs useful for lighting areas where you want to see clear colors on walls or furniture without interference from a tinted bulb. Daylight CFLs are stamped on the packaging with the word "daylight" and the specific color temperature they produce.

Then this:

Compact Fluorescent Photography Light Bulbs

it is from a commercial/vending site -
so I don't know what credence one can put on it.

However it is interesting that this "article" makes distinction between film and digital photography - citing film photography balances at 5500K - but digital balances at 6500K.

[fnj]

No matter what the CRI is of an artifical light source it can't replicate the spectrum of the sun, especially fluorescent tubes. Plasmas are close, while high CRI metal halide (we keep forgetting metal halide still exists) to my eyes replicates the sun better than any tube. If you wanted to replicate the sun's spectrum as close as possible in your living room, and coulnd't afford plasma you could always use high CRI halides in a torche' style fixture and would get really close. New ES series Rebels (5000k - 80/85CRI) are next on the list. The smooth spectrum of LEDs is something fluorescents can't match, but high CRI and high CCT is something LED's have only just started doing well.

Just received my BlueMax CFL and fluorescent desk lamp. 5500K, CRI > 93. Big difference in A-B test against even "daylight" and "sunshine" type CFLs. I am completely satisfied. Definitely matches or exceeds the best LEDs.