Would it suffice to say that sunlight is THE defacto basis for 100CRI? Even if other light sources fill in the “gaps” of sunlight spectrum, which apparently has gaps, it would be the same as a modern LCD supersaturating colors which aren’t actually as vivid as in person. The way sunlight blends does essentially fill in all the frequency gaps, I can’t think of a single color which doesn’t properly render under a high noon sun, physical or medical viewing limitations nonwithstanding.
If you shine a specific color light at something, it will either reflect, deflect, or absorb that frequency. If it does reflect that specific frequency, then there is some aspect of the pigmentation that reflects that color. If you’re throwing an artificially excessive amount of that frequency at the object, however, does the CRI actually go down, since the objective CRI basis is sunlight? Relatively, yes, because by comparison, all the other colors become washed out in relation to the new supersaturated color. Purple “doesn’t exist”, based on photometry. But by that standard, neither does white light.
People be making up all types of BS when it suits their marketing needs, and then claim that “the science shows”, when in fact it’s “our limited perception and observations show”. It’s why alot of our science has fallen behind, because the tech was specifically routed and directed a certain way.
Off topic rant over, it’s not specifically dialed at you, q/p, I just happened to tack on a peeve at the end…
Before anything else, I should clarify that the original post of mine you responded to was a based a misunderstanding on my part–I thought the light measured was the diffused light from the hazy sky rather than sunlight.
Excellent point in specifying the relevant conditions! There are many parameters to specify to ensure that you’re getting consistent sunlight: the “directly overhead” condition is only met along a strip of the Earth around the equator during a certain period of the year, with plenty of luck required for atmospheric conditions.
The 100CRI reference source is defined to be the blackbody source of corresponding CCT for 5000K and below, and standard daylight illuminants for higher CCT; in this sense, an incandescent lamp is equally valid as a “de facto” 100CRI source. However, the standard daylight illuminants are harder to work with for a few reasons, one being–as you mentioned–a collection of geographic and atmospheric conditions needed to ensure consistency.
Another difficulty is that while the blackbody spectra are a family of curves with closed-form expressions, indexed by a continuous parameter (temperature), the standard daylight illuminants enjoy neither property. They are derived not mathematically but empirically, and they don’t come in all CCTs: if you want a 100CRI reference in, say, 5243K, you are out of luck.
I believe this to be correct, and already observed in dedomed 519As: the R9 drops due to too much red, and R12 is universally low for high CCT emitters due to too much blue.
I’m not sure if I understood–would you elaborate a bit more?
Someone posted an article on how “purple” doesn’t actually exist as its own real wavelength, and that it only happens because our brains blend reds and blues together for some approximation of “purple”. And even in photometrics, purple “couldn’t possibly exist” because it takes the low AND high bands from visible light, while all the other colors are a specific frequency on the spectrum.
In short, purple exists is “an illusion”, but by that standard, so would white light, lol.
1: Yes, incandescent is 100 CRI 0.000 duv. Sunlight has some evident emission and absorption lines, it isn’t an ideal black body radiator, so, not 100 CRI and not 0.000 duv.
2: A deep red would be all R9, but some other bands will show some value too (R1 (faint red) will measure somewhat high)
I see. Saying that “purple does not exist” without any qualifiers is not only ambiguous but factually false–I recall a discussion thread on this as the title of a poorly-written article. Much better, as you stated, is to say that purple does not exist as a monochromatic, i.e., single-wavelength color. White light is a very nice counterexample!
There may be some distortions to the spectrum due to factors such as impurity of the glass, but an incan spectrum is as close to the ideal 100CRI as you can hope for, and certainly much, much closer than any LED approximation.
Loosely speaking, R9 measures the closeness between (a) the proportion of red in the test spectrum and (b) the proportion of red in the ideal spectrum. A red LED achieves a=1 while a decent white emitter should have b<0.5, not close at all. So a very low R9 should be expected.
Another way of seeing this: if all you have is a red LED, how reliably can you distinguish between subtly different colors close to red?
Ideal spectrum being sunglight or incandescent light?
Is it correct to say the flashlight with the highest R9 rating is the one that makes a red object look closest to the way it appears under sunlight or incandescent light? TIA
Both are ideal for their CCT–you need to first assign a CCT to your test spectrum before being able to compare it against an ideal source of equal CCT.
Thanks. Meaning incandescent light is an ideal spectrum only in its typical range of 2700k-3000k, whereas sunlight/daylight is an ideal spectrum at all CCT 's?
Almost. There does not exist a spectrum that is ideal for all CCT, not even sunlight. Instead, there exists a collection of ideal spectra indexed by CCT. A typical incan is ideal for 2700K, a typical halogen ideal for 3000K, and unfiltered sunlight ideal for 5700K.
Only spectra with the same CCT are comparable: to compare a 2700K test spectrum against a 5700K reference is meaningless: deviating from the 5700K reference does not imply low CRI of the 2700K source.
Hi there @QReciprocity42 and @cannga , a user found the CCT sub-discussion in this topic to be useful and suggested that it be moved into a separate thread for better discoverability. I can do that if you don’t mind?
Hi @sb56637! I’ve felt guilty about continuing this discussion in this thread, as it is not Convoy-specific; moving it to a separate thread would be perfect! Thank you very much for doing this.
“Numerically, the highest possible CIE Ra value is 100 and would only be given to a source whose spectrum is identical to the spectrum of daylight, very close to that of a black body.”
It’s interesting that per quote from wikipedia above, there seems to be an implied emphasis on “daylight,” as opposed to “sunlight.” I guess this means to see 100 score CRI, I would look and measure objects under cloudy sky or shaded area such as a patio (roof, but no surrounding wall)?
Daylight is the combination of all direct and indirect sunlight during the daytime. This includes direct sunlight, diffuse sky radiation, and (often) both of these reflected by Earth and terrestrial objects, like landforms and buildings. Sunlight scattered or reflected by astronomical objects is generally not considered daylight. Therefore, daylight excludes moonlight, despite it being reflected indirect sunlight.
