tint perception changes when the reference white balance changes
the blue and pink beams on the right side of the photos,
are from the exact same light, at different times of day
I give the sw45k a subjective personal preference score (TCR) of 10 during the day and 3 at night.
.
I have learned that the sw45k has a very high R9 CRI score, so I seek LEDs with similarly high R9 CRI scores (the red bar in the next images). The SST-20 also has high R9 CRI (above 80).
The sw45k has a very negative DUV score –0.0055.
I perceive this as Pink Tint, mainly when using daylight white as the reference
The SST-20 scores positive DUV of 0.0034, which to my eye is rather greenish, especially when observed from an incandescent white balance as a reference. I am not a fan of that LED’s Tint, despite the high R9 CRI spec…
I give the SST-20 4000k a subjective personal preference score (TCR) of 5 during the day and 0 at night.
I got 3 thermometers (clock/thermometer, clock/thermometer/hygrometer, thermometer/hygrometer), and despite being within inches of each other, always have a consistent 5°F spread.
So I have no idea other than relatively what the true temperature is.
Bought a “precision” glass candy thermometer (long-ass ones at least 1’ long), tried a stable water/ice mixture and boiling water, and even that seemed “off” by 1 grad each.
Same applies with voltmeters, lumenometers, etc. Unless they’re NIST traceable, they’re just educated guesses.
the DUV of sunlight is nominally 0.0032, almost identical to the SST-20 4000k charted above.
This is why that LED may not demonstrate a noticeable green tint, when daylight adapted
the DUV of incandescent is 0.0000, so from that reference, the Tint of sunlight is above the BBL, into the green zone
the LH351d and E21a also tend to have Tint above the BBL, in particular the cooler ones.
It makes sense to me that people that use lights during the day, will prefer Tints that are cooler and greener, more like daylight,
while people who use lights in darker ambient adaptation will appreciate warmer color temperatures and less green tints
sometimes it is useful to emulate sunlight, 5000k+
other times emulating incandescent can be useful 2700-3200k
I think it is helpful to match the color and tint of our flashlights to our ambient white balance
on the warm end, I like my artificial light to be about 500k cooler than ambient, so the beam stands out more easily. So when Im in 3000k ambient, I like a 3500k LED…
===
it occurred to me that the idea behind the TCR (what people find subjectively pleasing) is contained in the Kruithof Curve
that is, a group of people were asked what color temperatures they like at different Lux, and the results are plotted:
basically, at lower intensity, people like warmer light
“For example, natural daylight has a color temperature of 6500 K and an illuminance of about 104 to 105 lux. This color temperature–illuminance pair results in natural color rendition, but if viewed at a low illuminance, would appear bluish.
At typical indoor office illuminance levels of about 400 lux, pleasing color temperatures are lower (between 3000 and 6000 K),
and at typical home illuminance levels of about 75 lux, pleasing color temperatures are even lower (between 2400 and 2700 K). ”
What makes the rosy tint so popular? I guess we all agree that a flashlight with a duv greater than the duv of the ambient light doesn’t look as nice as one with a duv lower than that. But why?
Maybe that’s the reason, dunno. I asked because I wondered if there’s a paper about it. My first thought was that we’re used to incan lights, and so associate the lower duv with the comfort a child experienced at home. Something like this.
But yes, the higher reflection of red color makes skin look more healthy.
We don’t use flashlights the same way we use sunlight. We don’t use flashlights the same way we use lights in a house. Example: you are standing at one end of a 30 foot long white (walls,floor, ceiling) hallway in a building and you’ve got a light over your head and you’ve got a light that’s down at the end of the hallway 5 feet from the wall and you’ve got one in the middle all mounted on the ceiling. You want to see a painting on the wall down at the end of that hallway. It is a large painting of a deteriorating barn in a field with some blue sky. It is autumn and the field has not been cut. You have the option to turn on any single, double or triple combination of lights that you want. These are all bare bulbs shining in all directions. And you have three different bulbs that you can put in any position you want. One 2700k, one 4500k, and one 6500k. What bulbs do you put where and what do you turn on?
It seems to be due to both perception of what light looks “white” at different illumination levels and color temps, and that greenish/positive duv light often looks especially bad. I can’t really explain why it does to me either, but I’ve hated the light from greenish, low CRI florescent tubes all my life. Interestingly some florescent lights are very rosy instead, even when 70 CRI or lower. Keep an eye out for differences in temp/tint of florescent tubes right in the same or adjacent fixtures next time you’re around them.
I have one fluorescent tube left (nostalgia). My Opple measured a little over 2700K and a positive duv. It does not bother me at all, but yes, usually I prefer rosy over greenish tints. But I cannot really tell the difference between positive and negative duv. Yesterday I made experiments (not blind ones, obviously, I know my lights). With fully adapted eyes, I cannot identify the tint of light reflected from a white surface. Of course, there are limits. A sw45k D200 is outside of those limits. It is so much off that even my camera cannot find a correct WB. I’ve linked two reviews with some pictures in this thread, some of them show a rosy background. That was actually a white sheet of paper, and the light source was a sw45k, dedomed behind honeycomb TIR optics. I left it as it is since it was basically ok, but it was interesting to note that AWB didn’t work.
Ok, my idea about green/rosy is, when comparing two lights with different characteristics, one looks greenish, one looks rosy. This is really always the case in the NW range of CCTs. To me it shows the perceived tint is not necessarily tied to duv.
It is all a mystery.
Thanks for the paper and the presentation. The data is interesting by itself, even if it gives no hint about the psychology behind all this. I need to learn much more about it.
Edit: Maybe Whacky is not so wrong with his TCR. In some sense, at least. Interesting topic anyway.