When someone claims something they need to quote a study. Google is not a source is a search engine which will give you all the results you want and you don’t want.
That is why I have posted above a quotes from a study.
They don’t give any explanation for how the flame deviates from a black body spectrum. In my opinion that is incomplete, especially when they are attempting to use that 83 CRI number to show that their LED somehow gives off “better” light than a candle.
They don’t need to, but I always appreciate when someone is able to quote a study, especially if the claim does not match my expectation.
To get to the point, even sunlight does not perfectly replicate its own reference, which above 5000K is a standard illuminant that mixes direct sunlight and scattered skylight, in reality varies. I’d be somewhat curious to see measurements of sunlight at different times of day to see how measured CRI and DuV changes, and whether it ever quite measures 100, or tends to stay in the 99.x range.
Below 5000K, the reference for CRI is a theoretical black body, and there can again be some differences because of absorption and discrete emission effects that I assume are present. It should be extremely close 100, but I’ve also never seen an actual candle measurement.
I’m particularly curious about thorium mantle lanterns, because I know they have some strong discrete emissions, and might as a result be quite significantly short of 100 CRI.
Edit - just saw the paper linked above has a CRI measurement. I’m reading the paper now. 83 CRI surprises me and raises some questions, such as whether that is an integrated measurement.
That paper I suspect may not be accurate. We have people with spectrometers here on these forums that can run those tests easy to verify the claims. Mantle lanterns were already tested here. The mantle in question has a 98 CRI, essentially perfect.
Awesome! Thanks for doing this, and my compliments on the fancy white integrating toilet. It’s probably a good thing you didn’t try this one inside a styrofoam sphere.
98.4 is in the ballpark I expected, and the R9 is just about perfect. It is possible that the porcelain is not entirely neutral and is affecting the measurement of the reflected light, but I doubt it has a very large effect.
Nichia uses minimum. Some makers use average. Some users apply Nichia-style 4-digit specs to LEDs from other makers.
So it’s usually good to state whether it’s XXXX min. or XXXX typ.