Best Light for reading?

My personal preference is for the light to be whiter. If the light will be used only where the illuminance of what you are looking at exceeds

500-600 lux or so ("in my words half of slightly lowish side 'office-bright / classroom-bright'), I would use 4100K. That would be typically achieved within 25 inches of a 42 watt CFL or within 20 inches of a 26 watt one, or somewhat greater distances from a CFL in a fixture that has a reflector or at least a white surface behind the CFL.

If you will be illuminating your reading material less intensely than that or if you will also use this light for some of your general home lighting, my preference is to go 1 step lower to 3500 K. Higher color temperatures often have a "stark" or "dreary gray" effect with the lower illuminances common in homes.

There are linear fluorescents and CFLs of even higher color temperature ratings of 5000 and 6500 K. I find those to be even worse at "stark" / "dreary gray" effect than 4100. However, I have noticed that 5000-6500 K tends to achieve in most home use "an icy clean stark or dreary gray", while 4100 K deployed insufficiently to look "nice and bright" appears to me to be a "dusty-dirty stark or dreary gray".

--
 - Don Klipstein (don@misty.com)

I'd have to agree but also to add a bit of additional explanation: as noted, the "XXXX Kelvin" specification refers to the "color temperature" of the light source, which is roughly the temperature of a supposed "black body" radiator (an object which produces light solely through incandescence - think of the way a horseshoe glows red when taken from the forge). "Best" for reading, especially, is a matter of personal preference, as long as the source in question provides *sufficient* light for your needs (such that you're comfortable with it). The real concern with the various color temps (or "colors of white") has to do with how well they do at reproducing color accurately when you're dealing with reflected images such as the printed page. Incandescent lights, for instance, produce a MUCH "redder" light than, say, direct sunlight, so colors will look a good deal different in the two situations.

A "6500K" light source is supposed to be roughly equivalent to sunlight illumination (there are other conditions that go along with that, but that will do for now) and is fairly close to the "equal energy" white (a light source that's flat in terms of energy across the visible spectrum). Anything lower than that is basically biasing the white more toward the yellow, and eventually red, end of the spectrum, and will appear "warmer" (more yellow/orange/red) to the eye. Much above 6500K, and the light source is turning distinctly blue, and so appears "colder." Your eye adapts to anything that's reasonably "white," though, if it's the only (or primary) source in the field of view, so after a while you don't really notice that anything is wrong (unless you have some other source to compare colors to).

Bob M.

In article , Bob Myers wrote in part:

I do beg to differ, since sun's surface temperature and color temp. of direct sunlight outside Earth's atmosphere is more like 5800 K.

Overcast days appear to me to be less bluish than 6500K lamps - I have liking to consider "typical overcast conditions" to be 6000K even.

5500 is supposed to be some sort of typical of direct sunlight plus "to relevant extent" light reflected towards illuminmation of sunlight-iluminated photography subjects by "blue sky" and clouds.

I do sense some mention of to some extent using a "UV/Haze" filter to attenuate UV to an extent such that spectral response of "daylight color slide film" after modification by such a filter to result in a spectral response closer to that of human vision.

I do find "equal energy/power per unit wavelength version-of-white" to have "correlated color temperature" of 5455 K or by some accounts closer to 5400, maybe around 5420 or possibly as low as in the upper

5300's.

What about 4100 or for that matter even if that needs 4300-4400 K ("mildly overheated" "cool white" fluorescent lamp") appearing to be white? And "direct midday sunlight" in Philadelphia and nearby suburbs appears to me to achieve color temperature mostly 4400-4800 K, but "sometimes gets as high as 5100" as I see things here. I consider 5200 K to be an extreme of Washington DC on a favorably clear-air day close to "high noon"and close to "summer solstice", and 5400 K to be high side of "direct sunlight" from sun-at-zenith at lower altitude/elevation with low-side existence of whatever causes haze. As a result, I expect illumination onto a planar surface that direct sunlight is perpendicularly illuminating without obstructions to "light from ther sky" to average at 5500 K, maybe closer to 5775-5800 K in more-ideal situations of lack of cloud presence with sun at least

30-45-or-whatever degrees above horizon, or-similar... And "idealized 5500K daylight" still appears to me to be best-photographed in when the film is "color slide film" or "color movie film" uf a UV-attenuating filter is deployed in order to make the "roughly below-440-nm-spectral-response" of the film plus photographinh optics including filters more like that of human vision. That part does require attenuation of wavelengths near 400 nm and in the upper and mid-upper 300's of nm (including 350-360 and 390-390's).

--
Don Klipstein, best as I can say now, Best Regards to everyone/anyone, ..
(don@misty.com)

Yes, but therein lieth the problem - all of these identifications of various "color temperatures" (more correctly, "standard illuminants" like the CIE D65, etc., standards) as "daylight white" or whatever do come with a list of qualifications a mile long - things like "daylight white, with the scattering components removed, except on Thursdays in months with no "R" in the name, as measured at noon with a background of new-fallen snow." Stuff like that. Still, if you read a reference to "daylight white," it is VERY often the

6500K point or D65 illuminant they're talking about. For more on the gory details, see:

and

Note that actually, standard illuminants with CCTs ranging from about 5000K to 6774K are all referred to, with various qualifying statements, as "daylight white." The entire CIE "D" series was supposed to be "daylight" - but D65 (CCT of 6504K) is most commonly the one referred to as simply "daylight white." It may or may not appear bluish to various observers, depending on

- well, a whole lot of things.

Bob M.