I’ve been comparing the tints of 98CRI LEDs like Sunlike and Bridgelux Thrive vs. the tint of 95CRI 6500K LED (with blue spike) Yujiled. They seem to have a different appearance.
Sunlike and Bridgelux Thrive has a slightly “yellowish-greenish-white” appearance while the 95CRI with blue spike has a crispier "bluish-slightly-pinkish-white-appearance. The 95CRI with blue spike LED appears to have crispier cleaner white despite having a blue spike. sort of like looking on a white piece of pure white bond paper.
I don’t understand how a 95CRI standard LED with blue spike has a “cleaner” and “crispier” white compared to other violet pumped Sunlike with no blue spike or broad blue Bridgelux Thrive with no blue spike.
UPDATE 2
I compared other samples with negative dUVs
Sunlike 98CRI 7,407K dUV-0.006 - looks yellow-green compared to overcast daylight (overcast Daylight has CCT of 7200-7500K and positive dUV of about +0.002)
Yujiled 95CRI. 7,453K dUV-0.005 - looks pink compared to actual overcast daylight (overcast Daylight has CCT of 7200-7500K and positive dUV of about +0.002)
Yujiled 95CRI with plus green filter 7800K duv ,-0.0021 - looks bluish white with very small faint pink. almost similar tint to overcast daylight
the thing is, when compared to actual sunlight, using styrofoam as white target, the 95CRI led with blue spike is closer in tint to actual overcast daylight compared to sunlike or thrive.
I have the exact opposite finding! For the same CCT I find blue-pumped LEDs to be very pale, and violet-pumped LEDs to be slightly greener but much better balanced, and not at all pale. The greener tint is attributable to the removed blue spike (so less pink) and the filled-in cyan dip (so more green). The reduced blue and increased cyan, aside from rendering blues/greens more accurately, also make reds stand out much better; the difference is just as much as the difference going from R9050 to R9080.
Maybe your preference is just rosier or more blue-heavy tint, not the highest CRI.
I think you worded it pretty well!
I compared Sunlike with a few other blue-pump emitters. Not sure how to properly describe it - although the blue-pumps might have a rosier tint, the Sunlike makes colors feel more solid and rich, and it’s still has a tiny bit of rosiness.
Sunlike 98CRI 7,407K dUV-0.006 - looks yellow-green compared to overcast daylight (overcast Daylight has CCT of 7200-7500K and positive dUV of about +0.002)
Yujiled 95CRI. 7,453K dUV-0.005 - looks pink compared to actual overcast daylight (overcast Daylight has CCT of 7200-7500K and positive dUV of about +0.002)
Yujiled 95CRI with plus green filter 7800K duv ,-0.0021 - looks bluish white with very small faint pink. almost similar tint to overcast daylight
the thing is, when compared to actual sunlight, using styrofoam as white target, the 95CRI led with blue spike is closer in tint to actual overcast daylight compared to sunlike or thrive.
Do you have a picture of your comparison setup against sunlight? The fact that a 7407K emitter appears yellow-green relative to sunlight is very concerning, since sunlight is much warmer than this CCT and known to have +duv. I can think of 2 potential reasons why you are getting this:
Your sunlight sample might be contaminated by scattered illumination by the blue sky, which makes sunlight appear much cooler and more magenta.
Your styrofoam surface is fluorescent, which causes the UV in sunlight to react and produce blue, increasing the CCT and skewing the tint.
IMO a good way to tell the difference between blue and violet pumped emitters is looking at the palm of your hand. First look with a violet-pumped light, and observe that aside from red and cyan veins, there are also regions where the skin does not have a very definite color, noting how easily you can tell these non-colored patches apart. Now switch to blue-pumped, which makes these patches look yellow and blend in with the red veins (which also look less distinctly red), while the cyan veins are discolored.
Take an especially good look at the red bits–with blue contamination and not enough cyan to contrast, they don’t stand out as much.
Lol, looking at my palm has been my stable CRI check for a while. I wasn’t aware that it can even check for a violet pumped LED. You learning something new everyday!
Make a tight fist, then open and check the color of the finger tips. For a light source with good R9 you will see very obvious pink because of the blood being squeezed into the finger tips. With bad R9 the color is a lot more pale.
I’m not sure how reliable this check is for telling violet-pumped from blue–once in a while one comes across a rare blue-pumped LED that doesn’t have a blue spike and does have good cyan coverage. My Feit CCT-mixing light bulb has a 6500K setting that is not violet-pumped (looking through a spectroscope), but has no blue spike and much better cyan coverage than 519A/B35AM, and looks almost as good as my SunLike emitters.
In any case, it can reliably tell when (1) red or cyan is insufficient, or (2) there is way too much royal blue. Making a tight fist is a nice trick–I’m going to copy that. A high R9 emitter will make the reds look redder and more saturated than a low R9, while sunlight (and my SunLike 5700K) makes the red bits look almost hot pink! Probably because of the contrast from cyan.
sorry, what i meant was in comparison to “Daylight”, i compared it side by side with overcast daylight, no direct sunlight. The overcast Daylight has CCT of 7200-7500K and positive dUV of about +0.002.
I verified that the styrofoam wasn’t fluorescent since i exposed it to a Marktechopto 400nm LED. there is no bluish white glow just faint violet reflection.
my setup is very simple, just allow window daylight to shine on white styrofoam on the right side, then make a shadow on the left side. then i shine the Sunlike or Bridgelux Thrive on the shadow side making sure that there are no light spillover.
I see. I’m not sure what to say, 7500K is not really a common illuminant, and the SunLike LEDs (what is the nominal CCT?) is almost certainly not aimed to reproduce D75, so all guesses are off. If the LED is not rated as 7500K but produces a 7500K spectrum, there is something wrong with either the emitter or the secondary optical system. Plus, the spectrum of observed overcast daylight is affected significantly by many uncontrollable factors, such as extent of overcast, type of cloud, time of day, air quality, etc.
I can believe that the SunLike looks greener than the blue-pumped LED of the same CCT or even same duv, that is part of the design. Compare the emitters on blue/green objects (esp. light/sky blue) to see why some of us would prefer the SunLike over a blue-pumped light with less greenish tint.
By any chance, do you have color reports (CRI, Rf, etc.), or even better, spectrum plots of the emitters in question?
i only have CCT and dUV data from Lightmaster G2. i’ll update if i got a hold of Hopoocolor HPCS-310P spectrophotometer
Do you think if I got a Sunlike 6500K with a positive dUV or zero dUV, it would be less yellow-green in tint ?
I really like the richness of color of the Sunlike compared to a high CRI blue pumped LED but it is just that the “yellow-green” tint it produces that makes me not like it. In my eye, the high cri blue pumped led produces a somewhat “cleaner neutral white”. My eye cannot chromatically adapt to the Sunlike 6500K yellowish green white. My eye easily adopts to either purplish white or pinkish white making it look neutral white.
I’m excited to see your measurements once you get your hands on the spectrophotometer!
If you get a negative-duv SunLike, it would probably be less yellow-green. But these may be hard to come by because the spectrum is meant to mimic the reference (which is +duv) as closely as possible.
I think I understand your preferences. Objectively, the green SST20 4000K I have is much closer to afternoon sunlight than a neutral 4000K emitter, yet I prefer the neutral one despite the discrepancy. Also, the human eye tends to prefer pink sources in low intensities (and see neutral as yellow-green); if you could somehow crank the power of the SunLike all the way up to match the intensity of sunlight, you might end up preferring it over the blue-pumped one.
The adaptability of human vision is a tricky thing. If you add a blue spike to a smooth spectrum, in theory you could see it as higher CCT with more red than usual (which is pleasant), or around same CCT with more blue than usual (which is unpleasant). Maybe the way your eye white-balances happens to favor the former, and mine the latter. For me, any spectrum with a blue spike looks somehow simultaneously yellow and blue (which I experience as pale), and my eyes cannot balance to that unless adjusted to a lower-CRI source. Removing blue and adding cyan makes the overall tint greener, but eliminates the pale look.
Very interesting observations. Some matches my experiences. I, as many here, played with lights in the presence of afternoon sun light and found that the so called neutral DUV light appears pink in comparison to the actual sun. The 4000k sun light has a lot more green in it. It’s the human eye that auto balances out the green so it appears warm white. It also shows that color perception is largely subjective and relative. Being “sunlike” might not be a good thing in some situations.
4000K sunlight, given optimal air conditions, is extremely green, if I had to guess I’d say probably +0.01 duv. Outdoors it feels completely fine, but when shone through the window or used in low intensities (my SST20 has identical tint) makes everything look lemongrass green. Funny how human vision works!