Acebeam
Results 1 to 17 of 17

Thread: Pleasing light temperature VS illuminance level - Kruithof curve

  1. #1

    Arrow Pleasing light temperature VS illuminance level - Kruithof curve

    Very interesting wiki article: http://en.wikipedia.org/wiki/Kruithof_curve

    Knowing this, I find it very surprising that cool-white temperature LED's are popular at all. Everyone should be competing to buy the warm colors.

    QUOTE: "due to the Purkinje effect: the wavelength of light for which the eye is most sensitive depends on the luminance"

    QUOTE: "at typical home illuminance levels (75 lux), pleasing color temperatures are between 2400 and 2700 K"

  2. #2
    *Flashaholic* StarHalo's Avatar
    Join Date
    Dec 2007
    Location
    California Republic
    Posts
    7,439

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    It depends on where you are on your flashaholic journey; beginners start out wanting the brightest possible device that exists, then once they've blinded everyone in their house with a big gun light they realize something that lasts longer is more useful in most situations, so then begins the quest for low output lights, then comes the search for most useful interface, and then sometime around this point comes the search for most visually pleasing light.

    2400-2700K would be ridiculously orange for a flashlight, even our warm tint lights don't venture that low. 4000-5000K is the "neutral" sweet spot for most uses; my EDC is a very slightly warm ~4300K, and during power outages I use a cozy ~3500K light for living room lighting.

  3. #3
    Flashaholic*
    Join Date
    Jan 2004
    Location
    Pleasanton (Bay Area), CA, USA
    Posts
    3,593

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    It is useful to consider that different people have different goals, and equally important, the perception of color and CCT is not the same for everyone. The cited graph is a useful guide, but it is based on averaging perceptions together.

    Consider the simple case of a white shirt. People spend a lot of money on detergents that contain "whitening agents" (blue colors) to keep clothes from looking yellow. My wife has the same opinion about her kitchen, she wants a very clean, white look with no yellowish cast, so she likes high CCT, high spectral content lighting.
    Homebuilts - "BREEZE" RCR2 sidexside, "Tornado" 4 x 18650 side x side, Streamlight SL20x LED module (Custom BST Feeler Thread)

  4. #4
    Flashaholic*
    Join Date
    Dec 2007
    Location
    USA
    Posts
    1,604

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    I don't know if compact fluorescent bulbs use a different system for determining CCT [they shouldn't], because I purchased 4100K CFL, and it was more/less the same as 5000K LEDs. [it went back]
    3500K is a nice neutral CFL, and thank goodness people are demanding non-blue daylight, and they don't want CFLs to match incandescent.
    Last edited by MichaelW; 08-07-2012 at 12:19 PM.

  5. #5
    Flashaholic* AnAppleSnail's Avatar
    Join Date
    Aug 2009
    Location
    South Hill, VA
    Posts
    4,004

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by MichaelW View Post
    I don't know if compact fluorescent bulbs use a different system for determining CCT [they shouldn't], because I purchased 4100K CFL, and it was more/less the same as 5000K LEDs. [it went back]
    3500K is a nice neutral CFL, and thank goodness people are demanding non-blue daylight, and they don't want CFLs to match incandescent.
    Tungsten lamps are around 3200K 'color temperature' in most cameras. Daylight (direct sun, no sky) is 3000-4000K. Daylight (Sky only) is around 6000-7000K, while mixed sky and sun is around 5000K... but our eyes mostly perceive the sun's light, and pretty unconsciously correct for the mix of 3-4000 and 6-7000. After all, we've done it since birth!

    Interesting CCT note: I have a pet theory that the moon did not appear blue until the wide use of orange sodium-vapor lamps. During a power outage, or when far from such horrible glare sources, look at the moon. It is white at best, and greenish or yellow/orange. After all, it's only asphalt-colored rock in sunlight...
    My biggest light-hog is my camera.

  6. #6
    Flashaholic* kaichu dento's Avatar
    Join Date
    Apr 2008
    Location
    大和/Alyeska
    Posts
    4,930

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by AnAppleSnail View Post
    Tungsten lamps are around 3200K 'color temperature' in most cameras. Daylight (direct sun, no sky) is 3000-4000K. Daylight (Sky only) is around 6000-7000K, while mixed sky and sun is around 5000K... but our eyes mostly perceive the sun's light, and pretty unconsciously correct for the mix of 3-4000 and 6-7000. After all, we've done it since birth!

    Interesting CCT note: I have a pet theory that the moon did not appear blue until the wide use of orange sodium-vapor lamps. During a power outage, or when far from such horrible glare sources, look at the moon. It is white at best, and greenish or yellow/orange. After all, it's only asphalt-colored rock in sunlight...
    Great post here!

    I've long felt that discussions of daylight unfortunately left out the variables of angle in the sky, but to throw in the difference in direct sunlight and light coming from the sky makes it much harder yet for anyone to claim an ideal CCT for all. Diamond grading has long been done using 'north light', which is the light coming from the sky alone, without any direct light from the sun.
    Marduke - Solitaire...I've seen matches which are brighter AND have a longer runtime. 光陰矢の如し

  7. #7

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Big difference between CFLs and LEDs in regards to CCT: fluorescent light sources (and halide) have far more abrupt spikes in their spectrum than LED, which results in a much different aethestic even though CCT and CRI might be very similiar. Also, name brand warm white LEDs, while a much newer technology, tend to deliver aethestically superiour light in warm-white than fluorescent / CFL. Cree, Luxeon, Bridgelux etc., are pretty much delivering their best products in the warm-white market segment due to competition with perhaps Rebels ANSI being an odd exception. By an large though they *all* look exceptional for interiour lighting below 4000k. This is vastly different than the fluorescent and especially CFL market where some brands use whatever phosphor was scraped off the factory floor that week. Note that some brands of Fluorescent try to distance themselves from the low end market by specifying the phosphor set they are using.

    With the goofy lamps I'm building in fixed lighting all the neutral and warm-white emitters I've used (with the exception of Rebel ANSI or chinese LEDs) look stunning in a room at typical night illumination levels and far, FAR better than CFL.

  8. #8

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by AnAppleSnail View Post
    Tungsten lamps are around 3200K 'color temperature' in most cameras. Daylight (direct sun, no sky) is 3000-4000K. Daylight (Sky only) is around 6000-7000K, while mixed sky and sun is around 5000K... but our eyes mostly perceive the sun's light, and pretty unconsciously correct for the mix of 3-4000 and 6-7000. After all, we've done it since birth!

    Interesting CCT note: I have a pet theory that the moon did not appear blue until the wide use of orange sodium-vapor lamps. During a power outage, or when far from such horrible glare sources, look at the moon. It is white at best, and greenish or yellow/orange. After all, it's only asphalt-colored rock in sunlight...
    Actually, wrong. Pure sky (without sun) is in the 9-12000 range. Sun itself is a bit above 5000.

    I mean seriouly, make a sanity check on your statement. The SUN is not the same color temperature as am incan. Just take any incan light, go outside at noon and shine down at the sonlit floor. You will notice the incan to be A LOT more orange than the sunlight.

  9. #9

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    A while ago, I was considering installing HID headlights for my vehicle.
    I came across this web site which convinced me that I'm better off keeping my incandescent lights.
    http://www.danielsternlighting.com/t...dvantages.html

    I'm a fan of warm tints because they allow your eyes to be night adjusted, and therefore more useful for places where a flashlight is typically used.
    So you get less lumens, but consider that just 1 candle is enough to read a book with night adjusted eyes.

  10. #10
    Flashaholic* AnAppleSnail's Avatar
    Join Date
    Aug 2009
    Location
    South Hill, VA
    Posts
    4,004

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by IMSabbel View Post
    Actually, wrong. Pure sky (without sun) is in the 9-12000 range. Sun itself is a bit above 5000.

    I mean seriouly, make a sanity check on your statement. The SUN is not the same color temperature as am incan. Just take any incan light, go outside at noon and shine down at the sonlit floor. You will notice the incan to be A LOT more orange than the sunlight.
    I'd like to keep the tone polite and friendly. I was speaking towards cameras and perception, not physics. Here is one sanity checks about the CCT in cameras. My interpretation stands, for cameras and perception - not physics.

    Color temperature in your camera

    'Daylight' is sun + sky = medium CCT
    'Shade' is sky = high CCT
    'Sunset' is sun + (very little high-kelvin sky) = low CCT

    Technically speaking, the night sky is something like 10K white, only black. Unless you're near a city with skyglow. I am forced to wonder what the CRI of the sky is, since it is probably not a blackbody reflector at 10000 kelvin. The sun actually has a pretty decent CRI: What Color is the Sun?, but we see it here on Earth through some notable filters which alter its apparent CCT.
    My biggest light-hog is my camera.

  11. #11

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    For years working in the photographic industry has caused me to set mental standards on CCT values for outdoor lighting. The problem here is that the CCT of a clear day varies according to the height of the sun. The lower the sun the warmer the CCT. The effect though is not linear.

    Noonday sun at conventional elevations tends to hover around 6000-6500k. High powered strobes tend to run around 5500k. While it may seem an oxymoron to refer to the color temp of the sky without the sun it's actually quite reasonable because this is what occurs in the shade to some extent. Simply measure the CCT of a neutral target in the shade as long as the surroundings are reasonably neutral; pavement, snow, etc. This will tend to hover around 8000k, but is very sensitive to ambient surroundings, height of the sun and atmospheric clarity. Overcast sky tends to go cooler than noon-day shade, but never quite understood why this is.

    The comments about the moon are rather funny. The moon is actually the same color temp as the sun given the moon is pretty neutral in reflective albedo color and is essentially a large grey card in space. If you have a dSLR you can take a spot color balance of the full moon. Same rules for distance from the horizon apply. Moonlight only looks cooler due to reasons expressed in the beginning of this thread.

    For all intents and purposes moonlight, sunlight, electronic flash, incan etc.,should be assumed as 100 CRI sources.

  12. #12

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by AnAppleSnail View Post
    I'd like to keep the tone polite and friendly. I was speaking towards cameras and perception, not physics. Here is one sanity checks about the CCT in cameras. My interpretation stands, for cameras and perception - not physics.

    Color temperature in your camera

    'Daylight' is sun + sky = medium CCT
    'Shade' is sky = high CCT
    'Sunset' is sun + (very little high-kelvin sky) = low CCT
    During normal daytime (i.e. the sun reasonably high in the sky), there is almost an order of magnitude more direct sunlight than indirect, scattered illumination. Therefor, the 5500K of the sun (and the sun IS white, not yellow) dominate the total daylight spectrum.

  13. #13
    Flashaholic* AnAppleSnail's Avatar
    Join Date
    Aug 2009
    Location
    South Hill, VA
    Posts
    4,004

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    "The sun IS white, not yellow."

    That's the rub, though - "What is white?" In many situations, the objects around you determine 'white' far more than the altitude or time of day. And the light we see from the sun is always filtered. Finally, even out in space it's not a black-body emitter (Maybe because it's not a uniformly-composed emitting object, but I am unsure). We like to assume that black-body emitters do a good job of creating 'white' light, within reason (The 3 kelvin background radiation is not 'white.')

    What is white? 'An equal stimulation of all color receptors in the human eye,' or 'has no hue,' or something...but what is white? I can play tricks with 'white' light so that in a controlled, double-blind test, you'll pick tan swatches over colorless ones and proclaim them to be white as snow. White is perception, not a fact...and we can't even slap a wavelength limit on it as we do with red or blue.

    I agree that 5500K is an acceptable CCT for full-sun shooting, but I find 5500K CCT lighting sources to be about 1000K too cold for my liking. My absolute preference is to blend CCTs (3500-6000K CCT) in diffused indoor lighting, as this best simulates the blend of sources we get from sun and sky.
    My biggest light-hog is my camera.

  14. #14

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    I thought that article was mainly a bunch of hand waving but with very little "quality", i.e. referenced peer reviewed studies, scientific evaluations, etc.

    Warm tints allow your pupils to be unnaturally wide for a given amount of light. That has the effect of reducing visual acuity, depth of focus, and depth perception. I am not sure how any of those things are good for driving?

    They may allow your eyes to adjust, but they do not provide the wavelengths those adapted eyes can use, so what is the point?

    Semiman

  15. #15
    Flashaholic* kaichu dento's Avatar
    Join Date
    Apr 2008
    Location
    大和/Alyeska
    Posts
    4,930

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by SemiMan View Post
    Warm tints allow your pupils to be unnaturally wide for a given amount of light. That has the effect of reducing visual acuity, depth of focus, and depth perception. I am not sure how any of those things are good for driving?

    They may allow your eyes to adjust, but they do not provide the wavelengths those adapted eyes can use, so what is the point?
    For those who don't get the point, read the thread title again about pleasant light temps. It's worth repeating - there is no "best" tint for all - go with what you like and let others go their own way. You don't like warm tints, I don't like cold ones, but we can allow each other to choose what we like and still get along.
    Marduke - Solitaire...I've seen matches which are brighter AND have a longer runtime. 光陰矢の如し

  16. #16

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Kaichu, this was a comment on the article and the posts that have been made.

    - For my house, I have warm white in the home theater and in my bedrooms with neutral reading lights for visual acuity. The warm white is so that my body does not think it is daytime when I am going to bed.
    - In my kitchen I have neutralish (about 4000-4500K) and high CRI for visual acuity and because I can't get high CRI LEDS much beyond that color temp and I wanted LEDs for the effect/look I wanted.
    - Office, workspots --- home and work work are high CRI 5K for high visual acuity and proper circadian rhythm stimulation.

    When one is driving, the color of the light should not be chosen to make you happy, but to make you an effective driver. A higher CRI neutral\cool (but not cheap HID mod cool) is going to give you better visual acuity and make you a better driver by providing better acuity on the road and give you better peripheral vision.

    For the office, typically "warm" lights will not properly stimulate pupil response leading to eye strain and fatigue and if the work is precision, errors. None of those things makes you happy and again, circadian rhythm is important and warm lights are not good for that either. At work, happy does not equal --- feel like taking a rest.

    In your home, whatever makes you happy (just don't say incandescent please, the planet needs your help) is the way to go!

    Semiman

  17. #17
    Flashaholic* kaichu dento's Avatar
    Join Date
    Apr 2008
    Location
    大和/Alyeska
    Posts
    4,930

    Default Re: Pleasing light temperature VS illuminance level - Kruithof curve

    Quote Originally Posted by SemiMan View Post
    Kaichu, this was a comment on the article and the posts that have been made.

    - For my house, I have warm white in the home theater and in my bedrooms with neutral reading lights for visual acuity. The warm white is so that my body does not think it is daytime when I am going to bed.
    - In my kitchen I have neutralish (about 4000-4500K) and high CRI for visual acuity and because I can't get high CRI LEDS much beyond that color temp and I wanted LEDs for the effect/look I wanted.
    - Office, workspots --- home and work work are high CRI 5K for high visual acuity and proper circadian rhythm stimulation.

    When one is driving, the color of the light should not be chosen to make you happy, but to make you an effective driver. A higher CRI neutral\cool (but not cheap HID mod cool) is going to give you better visual acuity and make you a better driver by providing better acuity on the road and give you better peripheral vision.

    For the office, typically "warm" lights will not properly stimulate pupil response leading to eye strain and fatigue and if the work is precision, errors. None of those things makes you happy and again, circadian rhythm is important and warm lights are not good for that either. At work, happy does not equal --- feel like taking a rest.

    In your home, whatever makes you happy (just don't say incandescent please, the planet needs your help) is the way to go!

    Semiman
    As mine was towards post which suggested that the criteria that you use for choosing a product should be what all others do as well. Too much government meddling with personal choice already and I don't think we should be doing that here either.

    I like the reasoning you have for your choices, but won't tell others they can't choose differently. Love my incandescents and look forward to them being bested in the market place by better and better performing LED's.
    Marduke - Solitaire...I've seen matches which are brighter AND have a longer runtime. 光陰矢の如し

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •