[Review] SST-20 in FW3A is divine!

Bob_McBob · 2019-09-20T03:22:32.000Z

Before commenting any further, everyone should read this article carefully. It explains the NIST and other research into colour perception and preference TK and I are discussing and how it has influenced ANSI colour standards. Not only were very negative Duvs strongly preferred at all CCTs by most test subjects, but the cutoff for even perceiving a tint as white is also much lower at 4000K and below. For most people, a perfectly neutral Duv is not preferable after becoming accustomed to the light source. That doesn’t mean there shouldn’t be any neutral lighting, or that a daylight reference isn’t useful, but user preference is ultimately what matters.

https://www.ledsmagazine.com/smart-lighting-iot/white-point-tuning/article/16695905/ansi-continues-advancements-on-ssl-chromaticity-standard-magazine

Joshk · 2019-09-20T03:25:38.000Z

I think we posted at the same time Bob_McBob. Yea I could believe that your vision could shift that far in a completely dark room. But daylight blends in from everywhere in our daily lives.

contactcr · 2019-09-20T03:38:08.000Z

These two experiments on 18 and 21 people appear to just be a way of requesting further funding and research for a more comprehensive study. While I don’t doubt people prefer negative Duv (to what extent who knows) this article is far from gospel and highlights that people like TK are still “not average” and apparently neither are people who prefer something closer to the BBL.

ToyKeeper · 2019-09-20T03:43:37.000Z

I understand that it can be frustrating when things don’t quite line up. Humans evolved on this planet, so human vision should be perfectly aligned with this planet’s natural light source. But it’s not… and that’s okay. Almost nothing ever lines up perfectly. The universe is an incomprehensibly vast gallery of slightly-tilted paintings.

SKV89 · 2019-09-20T03:48:04.000Z

In the past, I’ve asked many non flashlight folks to get their opinion on which beam color they prefer the most. If the CCT is about the same, EVERYONE ALWAYS say they like the below BBL tint more than On the BBL and they dislike the above BBL tint. Sunlight is above the BBL and to me only looks great only when the intensity is high and does not look good when only a faint amount of daylight gets in the house. However, no flashlight can achieve lumen output anywhere close to sunlight so for most cases during night time, I can confidently say the majority will prefer negative duv.

My ideal is slightly rosy with DUV between –0.002 to –0.005. 219B SW45K measures –0.01x on turbo, which is a bit too pink but I still prefer it over a positive BBL. In fact everyone I showed my 219B sw45k flashlights to like the tint better than any on or above bbl 4000K-5000K flashlights.

CLB · 2019-09-20T04:58:42.000Z

Bob_McBob:

very negative Duvs strongly preferred at all CCTs by most test subjects,

Why, exactly, are we invoking merely statistical studies here, especially ones with all of sixteen people as subjects? If these studies are meaningful, it is to lighting marketers, not consumers. As the researchers clearly point out, “Color preference or perceived color is subjective. It depends on the light level, the surrounding environment, or even the observer’s cultural background.” So in what way are preference statistics relevant to what you or I like, or to the relative chromaticity merits of the SST-20 versus the 219B (which is how this discussion started)?

Edited to remove my own confusion between “perceived whiteness” and “preferrred tint.”

Joshk · 2019-09-20T04:30:40.000Z

Ok I stepped away to take some pics.

If we light the right side with “daylight” and the left with rosy, our eyes need to pick a white balance. They pick daylight, and make the left side red-tinted.

Now if our brain sees the rosy side as white, but there is daylight too, the daylight looks green. And again this is weird.

But if we have matching light sources, there is only one white point and all is white.

So unless you can eliminate the daylight from the sky, you are forced to agree with it.
And yes, if you do eliminate the daylight, there is a preference for a rosy lights, but you don’t really notice it’s rosy after you adjust.

CLB · 2019-09-20T04:43:41.000Z

ToyKeeper:

I understand that it can be frustrating when things don’t quite line up. Humans evolved on this planet, so human vision should be perfectly aligned with this planet’s natural light source. But it’s not… and that’s okay. Almost nothing ever lines up perfectly. The universe is a vast gallery of slightly-tilted paintings.

That’s hanging a lot of metaphysical weight on a couple of very small studies. Generally, when things appear not to line up, scientists look for a reason. And they usually find it (or hypothesize it). Are you sure the paintings are tilted?

Bob_McBob · 2019-09-20T05:00:11.000Z

CLB:

That’s not how I read the results:

The figure you posted is from the LRC research on what people perceive as untinted white light at various CCTs. As you can see, this is slightly above BBL at higher CCTs, but quite a bit below BBL at lower CCTs. The NIST research further demonstrates that regardless of CCT and perception of neutral tint, most people still prefer a very negative Duv at all CCTs. Further research has broken this down more, but it is clear neutral light is often not actually preferred at lower CCTs in particular even when carefully controlling for numerous other factors in ways that wouldn’t necessarily be feasible in general lighting, and negative Duv often naturally correlates with these as demonstrated in previous research.

https://www.energy.gov/sites/prod/files/2018/06/f52/RoyerWilkersonWei_SageJournals_August2017_1.pdf

Bob_McBob · 2019-09-20T05:05:48.000Z

Joshk:

So unless you can eliminate the daylight from the sky, you are forced to agree with it.
And yes, if you do eliminate the daylight, there is a preference for a rosy lights, but you don’t really notice it’s rosy after you adjust.

There is a preference for rosy tints at 2700K even after adjustment and careful experimental control of colour rendering in the reseach I posted. You keep talking about daylight, but in the vast majority of cases where I use artificial lighting, it’s either the primary light source or it’s after dark. You can’t just say the sun always comes first in colour preference and only 12 hours of the day actually matter to most people.

Joshk · 2019-09-20T05:09:39.000Z

Yea, I agree with everything you are saying Bob. I never said daylight is first in preference. I was saying that dang thing didn’t come with an off switch.

ToyKeeper · 2019-09-20T08:13:15.000Z

CLB:

in what way are preference statistics relevant to … how this discussion started)?

I think it was mostly because of comments like this, which don’t seem congruent with empirical data:

BlueSwordM:

The SST-20 is better than the 219B in every aspect, even in terms of tint

Joshk:

the 219B is obsolete by all perceivable measures.

Some comments even seem to make fun of those with different preferences:

Joshk:

green-tint phobia

contactcr:

I love these new below BBL LEDs for photography!!!!

I haven’t been particularly nice either. I tend to call out errors and jabs which are probably too minor to really justify attention, and some of my comments probably sound pretty condescending.

… and after starting, the arguments continued when people doubled down instead of updating beliefs to match the data.

To be clear, SST-20 isn’t bad. In some cases, like for some members of this thread, it’s pretty close to optimal. And that’s awesome. But that’s a somewhat atypical result, and the “obsolete” emitter seems to fit the most common case a bit better so it’s inaccurate to call it obsolete.

Of course, none of that matters for personal use. Only individual preference matters there. It’s just a bit misleading to present personal preferences as if they are true for everyone. We could avoid most of this just by weakening the wording a bit, like saying “219B is obsolete for me” instead of asserting it’s obsolete period.

But the disagreements continue because exaggerations continue:

Joshk:

for the majority of use cases there will be daylight bleeding in

Joshk:

[My eyes] pick daylight, and … you are forced to agree with it.

Due to the strength of the assertions made, framing one person’s perceptions as normal and others as aberrant or nonexistent, it sounds as if being in the pink-biased part of the population is somehow bad, like there’s something wrong with people if they don’t prefer 0.000 duv to all other tints. Those people are implied to be invalid or broken. Maybe that’s not the message it was intended to convey, but if not, it would be helpful to word things a bit less aggressively.

Anyway, I hope this episode of “Today’s Argument, Abridged” answers the question adequately.

CLB · 2019-09-20T05:22:23.000Z

Bob_McBob:

The figure you posted is from the LRC research on what people perceive as untinted white light at various CCTs. As you can see, this is slightly above BBL at higher CCTs, but quite a bit below BBL at lower CCTs. The NIST research further demonstrates that regardless of CCT and perception of neutral tint, most people still prefer a very negative Duv at all CCTs. Further research has broken this down more, but it is clear neutral light is often not actually preferred at lower CCTs in particular even when carefully controlling for numerous other factors in ways that wouldn’t necessarily be feasible in general lighting, and negative Duv often naturally correlates with these as demonstrated in previous research.

You’re absolutely right about my confusion between “perceived whiteness” and “preferred tint,” which I realized as soon as I re-read the article, and I have edited my earlier post. Thanks. But I still am not seeing the relevance of these studies to the discussion of the chromaticity merits of two different emitters. If we were discussing the relative literary merits of two books, popularity would be a pretty uninteresting metric. In books, food, and light, I suppose what is popular is not identical to what is “good” from some critical standpoint or another. Now, the question of WHY people would consciously prefer a tint which is perceived as “less white” is interesting.

iamlucky13 · 2019-09-20T05:27:30.000Z

DavidEF:

Joshk:

Ah ha! It’s actually a Locus that merges the BBL and the Daylight Locus. Very cool. It does so by using the 5000k-4000k region to transition.

But… WHY? If the Daylight Locus is more green than the BBL, what is the supposed advantage of “merging” them this way? To say it another way: What light source is anyone measuring that they’d want to skew the “results” in this way?

The most common reference illuminants use different typical hues of daylight for color temperatures of 5000K and above, and the BBL at lower color temperatures.

There is already a disconnect in the typical color temperature ranges of typical light sources - obviously daylight because the sun is the most prominent one, and traditional incandescent bulbs at the BBL. I’d say it makes more sense to bridge that with a gradual transition than a sudden jump.

I’ll stay out of the discussion about duv preferences other than to note reasons to choose one way or the other vary depending on what you are trying to achieve. Since if I remember right, Josh is a photographer, I expect his priority to be producing a good match with the ambient light he has to work with.

I haven’t done much photography, but I was pleasantly surprised to see how well it actually worked for the sake of the photos the first time I made the light from my flash “worse” in terms of typical BLF preferences, but adding a tough plus green filter to better match the fluorescent lights in the building I was taking some event photos in.

SKV89 · 2019-09-20T05:27:33.000Z

Joshk:

And yes, if you do eliminate the daylight, there is a preference for a rosy lights, but you don’t really notice it’s rosy after you adjust.

I’m not going to debate with you regarding flashlight usage during the day when there’s daylight because I only use my flashlights in the dark.

However, I would like to disagree with your point that we don’t have a preference once our eyes are adjusted to a light source. My wife and I have been complaining for a few years regarding how lime green our in door lighting is, especially in our bedroom where we have a slightly yellow toned paint. Even after hours sitting in the room, it still bothers us and my wife said so many times that we picked the wrong color paint. I’ve spent a few months buying and testing dozens of light bulbs and led strips to replace the light bulbs and led strips in my house. I have led strips in all my rooms. I even bought the Sekonic C-800-U to assist me in finding good quality lighting. I found out the old 3000K led strips had DUV of 0.006 to 0.007 whereas the light bulbs were 2700K to 3000K with DUV of 0.002 to 0.004. The biggest difference I experienced was with replacing the led strip in the bedroom with a 2700K 95CRI led with –0.005 duv. OMG the difference was huge, despite the warmer temperature everything just looked so much more natural and whites were white (after eyes adjust to the cct) and no longer permanently lime yellow. Yes you are right after a few hours in the room we don’t really notice it is rosy but everything just feels so much more comfortable to look at, whereas with the old led strip no matter how many hours, it still looks yellow/green and uncomfortable. We no longer think we picked the wrong color paint for the bedroom wall. With the light bulb replacements the difference weren’t as big but we definitely can feel the rosier bulb enhanced the appearance and comfort during nighttime.

SKV89 · 2019-09-20T05:43:12.000Z

CLB:

Bob_McBob:

very negative Duvs strongly preferred at all CCTs by most test subjects,

As the researchers clearly point out, “Color preference or perceived color is subjective. It depends on the light level, the surrounding environment, or even the observer’s cultural background.

From my experience showing lighting to non flashlight folks, the preference so far has always been completely towards negative DUV. However, with CCT, I hear different preferences. Some like cooler temps while some like warmer temps but it also depends on the time of the day I show them the comparisons and what lighting their eyes are already adjusted to. When I show the lights in the day time for comparison, most people prefer the NW/CW CCTs whereas at night time, most prefer the WW/NW CCT.

Cultural background should also be a major factor. A Chinese contractor suggested to me to buy CW lights for my house remodel because he said it makes everything look so much cleaner instead of ugly yellow and you feel so much more awake with white light. I explained to him I prefer WW because it looks cozier and I rather feel relaxed and sleepy at night. He told me in China, everyone likes CW light but only in the US does he see so many people use WW. So I think that is the main reason why so many flashlights come in CW only because the flashlight makers have a personal preference for CW and probably think the entire market prefers CW also.

djozz · 2019-09-20T08:19:55.000Z

Nice discussion, I learned a few things

Joshk · 2019-09-20T12:49:58.000Z

This will likely be my last post on the topic as I’ve grown tired of everyone doubling down on ‘preferences’.

White point - Is a point, you can only have one. You must only have one, or there will be weirdness. Ask a photographer.
Rosy preference- Yes under lab conditions where there is only one WHITE POINT, people prefer rosy. Of course calling it rosy implies you are now comparing it to a second white point.
my ‘obsolete’ comment - That was a bit strong, how about this: “If brightness matters at all to you, the 219b is not the best choice.”
The TM-30-18 locus - This isn’t a flashlight world thing, it’s a real world thing. That’s why it cooperates with daylight.

mbp · 2019-09-20T13:48:44.000Z

Bob_McBob:

CLB:

That’s not how I read the results:

The figure you posted is from the LRC research on what people perceive as untinted white light at various CCTs. As you can see, this is slightly above BBL at higher CCTs, but quite a bit below BBL at lower CCTs. The NIST research further demonstrates that regardless of CCT and perception of neutral tint, most people still prefer a very negative Duv at all CCTs. Further research has broken this down more, but it is clear neutral light is often not actually preferred at lower CCTs in particular even when carefully controlling for numerous other factors in ways that wouldn’t necessarily be feasible in general lighting, and negative Duv often naturally correlates with these as demonstrated in previous research.

https://www.energy.gov/sites/prod/files/2018/06/f52/RoyerWilkersonWei_SageJournals_August2017_1.pdf

Correct me if I’m wrong but I don’t read the study you cited as supporting your point that “most people still prefer a very negative Duv at all CCTs.”

Page 16 of the DOE study: “The simple interpretation from a visual examination of Figures 4 through 7, which show mean data for each color rendition condition and for each chromaticity group, is that Duv has an effect on perceived normalness and preference at lower CCTs (2700 K), with negative Duv values being perceived as more normal and more preferred. The effect of Duv is negligible at 4300 K.” I don’t think I’m taking this quote out of context as it’s repeated almost verbatim in the final conclusions of the article. The study overall seems to conclude that Duv and CCT are not the most useful variables in predicting consumer light quality preferences: “CCT and Duv should not be used alone as factors in a model of color preference; adding both offers a small increase in predictive value for IES TM-30-15-based models of color preference.”

The LRC study also, as you stated, found that participants found positive-Duv to more closely resemble “untinted white light” above 4000k; negative-Duv below 4000k. The NIST research does appear to track more closely with your claim, but that’s 1 among 3 studies—so the weight of your own evidence is moreso against your claim, as I read it. Unless you meant strictly “most people prefer negative Duv at 2700k/incandescent light levels.”

Bob_McBob · 2019-09-20T15:35:22.000Z

The LRC study was not testing preference, just perception of neutral white tint, and the results are pretty much replicated in the third study. The new study attempts to control for colour rendering factors typically associated with negative Duv to demonstrate it isn’t the chromacity itself influencing preference, but even then there was still a clear preference at one of two comparison CCTs. This may indicate that at higher CCTs, fine control of specific colour rendering could eliminate a preference for negative Duv, but since negative Duv is naturally associated with these factors (based on other discussion of this research), it doesn’t necessarily follow that the preference could easily be eliminated in many general lighting situations, e.g. most consumer lighting products.