High CRI... not what I expected? (Happy ending in post 118)

I normally see four of everything--that's normal, right?--but now I'm seeing sixteen women in that pic!

All the pretty colors!

(I think I need to see my optometrist.)

Me, too. My pessimetrist is just getting me depressed.

[quote=twisted raven]

[quote=jon_slider]

This is the most important point of the discussion and I wish it was highlighted more often. CRI is a comparison to a blackbody OF THE SAME CCT. As soon as you change the CCT the comparison is largely meaningless. To say I like a high CRI 4000K over a low CRI 6500K says far more about what CCT I prefer than the CRI.

You’re welcome to disagree with every word stated, however, that doesn’t alter the fact that CRI is a comparison with a black body OF THE SAME CCT. A very warm CCT (let’s say 2700K) that has particular CRI won’t automatically render colours more accurately than a cooler CCT with a lower CRI. The CRI value is actually meaningless without knowing what the colour temp is.

I’m not saying you can’t enjoy higher CRI lights but to simply say “high CRI is the most important aspect of a light” is leaving out the more important part of the equation; what colour temperature do you like?

I’ve actually been trying to follow this thread and understand what each person is trying to say ( and by that i mean i’ve been goggling the terminology ). I’ve goggled that stuff before but it got so scientific and convoluted that i lost interest. I think i have a better understanding of bbl and duv now but i still don’t understand some people’s comments.

If i understand this correctly CRI is colour as viewed in midday sunlight? So any cct other than 5500K is just going to be …err… wrong. Before you jump on me to argue that point, I came to this resolution while researching then I found this website that said exactly that. Sure the bbl is to accurately show black … at a given temperature (high CRI). That doesn’t mean colours will show exactly as seen in sunlight or 5500 high CRI (because our eyes adjust to what they are seeing…?). On top of that we have each individuals perception of light and on top of that each has their own preference of what feel comfortable in seeing. The comment i made about our eyes adjusting is just something else on top of all the other variables.

I fall in the 8% of males with colour blindness - I see 21. Some argue that percentage should be much higher. I’ve known about this colour blindness since i was a teenager and only had 3 maybe 4 occasions in my life (45) when someone has pointed out to me that i wasn’t seeing the colour quite right. From my understanding/experience the blindness sometimes depends on the surrounding colours/light. Hence i’m half interested in the poll that BurningPlaydoh suggested.

[quote=Dickens]

[quote=twisted raven]

After looking up this stuff I think the above is the safest/best way of finding a starting point for ones personal preferences (cct) before experimenting in real world tint and CRI. From my experience - real world experimentation is key (insert specific application if needed) and i wonder how many hobbyists actually find their preference when they are comparing 5500K low CRI to 3000K high CRI.

I think we all see differences in cri regardless of colour temp or tint. Though, I think the purpose of that chook picture is an unfair comparison for the reasons mentioned by twisted raven. BTW the pink chook looks uncooked to me. Though i am interested to know why beam geometry makes a difference…? Is this related to tint shift within the beam?

I totally agree with your point. With the choice of leds available nowadays it’s (for lack of a better word) a minefield to find ones true preferences albeit just one aspect (big or small) of the hobby.

Yes, CRI is a vague objective description of how accurate a particular light is compared to an ideal black body radiant of the same CT.

As an example, one can’t just say ‘x led at 5500k with 99 R9 has more deep red output than y led at 3000k with 80 R9.’ 5500k by its nature has way less red content to it than 3000k, so an led that is 99% accurate at deep red output at 5500k might still have a fraction of the deep red output of a 3000k that is 90% accurate at deep red output.

Can you? How easily can you pick out the CRI differences in these samples, where exposure, CCT and beam geometry have been somewhat normalized? (tint differences not normalized) One LED has a CRI of 65, and the other a CRI of 80:

In this example, one led has a CRI of 99, and the other a CRI of 80:

For home use, I think CRI is probably the most important aspect of an LED light bulb.

to those who are going off into CRI theory, I think youre missing the point

I prefer the High CRI LEDs I actually have, to the Low CRI LEDs I actually have…

this is not theoretical

I KNOW I like my High CRI LEDs better than the Low CRI ones. I believe that is because my High CRI LEDs are better at showing Reds. This is not theoretical, it is my experience, and I have posted photos to demonstrate and explain. The photos were not doctored, I did not alter the image in post production in any way.

my point is, to those who choose to disagree with my opinion, post some real photos of real LEDs and demonstrate your opinion that it is not the High CRI that is making me happy. If you think it is just the tint or CCT, show pics to support your opinion.

Either way, whether the reason I prefer High CRI is due to R9, Tint, CCT, or hot spot size… there is no denying that for me, the sw45k is better than a NW Low CRI XM-L2, even when the CCT is the same…

anyway, enjoy whatever floats your boat

for those trying to wrap their head around Tint, aka DUV, aka above or below BBL… here is a visual (image thanks to maukka):

the BBL is the 0 DUV line
look how all the 219c have tint above the BBL
look at the sw45k, see that it is below the BBL, aka, has negative DUV

I do have my own color temperature preferences, but no matter what CCT, I will always choose the High CRI option with the highest R9 and the least green tint…

all the variables matter… it is not a One Variable is Most important, type of situation. But still High CRI is an absolute requirement for me. I will not trade it for CCT nor Tint. I want it all… the CCT I like, the Tint I like, AND High CRI…

I have an EagTac D25A/Nichia 219B sw45k for indoor and a TrustFire T62/XHP70/6000K IINM for outdoor with me this weekend. Both do an excellent job for what I want them.

Ha, did a medical inspection (wood splinter) with the D25A, that was so cool :+1: .

The way I understand that is cri reigns supreme and you can easily see the difference like 2-26350-P91 can. And you don’t need side by side comparisons to see the difference. So when out camping/ walking the dog or whatever a 6500K high cri is more comfortable on the eyes than a nice warm low cri.

Personally >5000K is a bit stark for me so Kelvin reigns supreme.

CW can be warm too. Under a 10000K sky, it looks pretty good. Without any environmental light, it’s ok, too. The eyes do what a camera does with white balancing.

Yes, yes, trash camera, trash editing, trash everything. But amazingly reproduces what I saw.

In this context: Is there a $40 cam that is usable?

Multiple scientific resources please. Otherwise, you’re free to your opinion.

No, definitely not to others or most. Higher CCT white lights with greater blue content will generate worse glare that people will falsely interpret as the light being “bright”, see Jim Benya’s take on the matter. Then there was also this research by UMTRI that showed that higher CCT LED headlights generated worse glare then lower CCT headlights.

terminology:
CCT = Corrected Color Temperature, expressed as Kelvin… eg…4000k, 5000k…
DUV = Tint, it plots the LED Tint in relation to the Neutral reference, aka bbl, Black Body Line
CRI = Color Rendering Index. measures the percent of a given spectrum output. Red Spectrum, aka R9 CRI. On High CRI LEDs, R9 is a positive number, on low CRI LEDs R9 is a negative number.

R9 is a CRI factor, it is not what we see as the Tint
Common usage of the word Tint, often is misapplied to mean CCT…

Daylight White Balance = 5600k
Incandescent White Balance = 2700k

White Balance of the brain is a primary factor affecting how warm or cool an LED “looks”, at the time.

For example
If I have a 3000K 219B, and I wake up in the dark, it does not look orange, nor “too warm”, at that time.

otoh, the same 3000k during the day, when my brain is Daylight White adapted, makes the 3000k look extra orange, and “too warm”.

Because white balance affects comparisons, I try to use a consistent daylight white balance for photos.

but in actual use at night, a 3000k light does not look as orange as it does during the day.

here is an example of three different Tints, CCT, and R9 CRI

the most apparent difference is the Tint, followed by Color Temperature.

The most subtle difference is R9 CRI. The above photo does not show R9 CRI at all. To see R9, the light must be illuminating a Red object…

Good observations Jon_Slider

This is why I wish there would be more variable color temp lights.

The optimum color temperature for looking at something in the daytime outside is quite a bit cooler than at night with dark adapted eyes. I think an Emisar D4 with variable tint like the Sofirm IF25, and running Anduril 2 would be lovely.

When choosing an LED for my lights I tend to consider the following in order of priority:

  • Output - A flashlight can have the best CRI, tint and color temp in the world, but if it doesn’t actually produce enough light to illuminate what I want to see what’s the point? I like pocket rockets so more output is always better. But that said, once the output is “good enough”, then other factors such as CRI become more important.
  • CCT - if the color temperature is too cool, I’m probably not going to like it no matter what other features the LED has. Since most lights aren’t variable tint, I have to pick a compromise that looks good to me in most situations. My preference is 4000-4500K, with occasional use up to 5000K. I tend to not use lights with other color temps no matter what features they have.
  • Intensity - Closely related to output. More intense LEDs throw better. Depending on the optic or reflector your light uses this could be a good or bad thing. Also, in some small lights I can get away with a less bright LED so long as the hotspot is still bright enough. For dedicated throwers this is the most important factor.
  • CRI - after all the others are considered, I prefer high-CRI with high R9, such as the SST-20 4000K. But for me, CRI is probably the least important of the above factors. I don’t like Nichia 219B sw45k, because every light I tried them in failed to produce sufficient lumens to satisfy me. On the other hand, I like SST-20, because they have a good mix of lumens, intensity and CRI, despite having less overall output and intensity than XPL HI (with both at 4000K).

I agree :slight_smile:

my lumen requirements are at HDS levels… 2-300 lumens is seldom needed, for the way I use my lights at indoor distances and close range outdoors

Tint is a high priority to me, so is Color temperature
the conclusion Ive reached, is that even after all those criteria are met, I will choose the LED with the best CRI, and for me, the most limiting factor on CRI, is R9, Red Spectrum

so, I can agree with others who say lumens is their first priority, or tint, or CCT… but after the first priority, there are a few others that need to Also be part of the Total Package.

I want all of it, Output, and Tint, and CCT, And CRI :+1:

thanks all for the discussion and exploration
enjoy your choices!

Ah, yes I remember the first time I used 3000K during the day and thought “wtf is with this weird orange colour…?” I knew it was meant to be a warm temp but wasn’t expecting the funky colours it generated to my daytime eyes.

Which makes me wonder how much the brain can counter for all the variables. The pink BBQ chicken looks uncooked but if one were cooking using ONLY that light then our brain would adjust and the chicken would look more appetising. As indicated perception of light largely depends on environmental factors (other light sources at the time). So I still have a hard time understanding why “high CRI” generates so much interest. My guess is unless out camping or the like our eyes are adjusted to home lighting in the evenings and that’s the base point from where our brain makes the ‘balance’.

The brain is very good at adjusting its white balance to match the CCT of a single ambient light source.

The brain takes about 30 minutes to downregulate from daylight cool white, to incandescent warm white.

The brain definitely adjusts to make 3000k the “normal” white reference, IF that is the ambient CCT of the home lighting.

During the day, the brain is white balanced above 5000k… due to daylight.

in both cases, the ambient light, creates a setpoint for the brain’s white balance reference.

From there, turning on a second light source, an LED flashlight, will reveal any difference between the brain setpoint, and the LED.

IF the LED is warmer than the brain white balance at the time, the LED will look warm, or orange

this is the case when using a 3000k LED during the day.
otoh, a 5000k LED will not look cool or blueish, when the brain is adapted to daylight.

IF the LED is cooler than the brain setpoint,
for example, under incandescent house lights, a 5000k LED will look cool and blueish.
but the 3000k LED will not look very orange when the brain is adapted to warm light…

warm light is more relaxing
cool light is more stimulating

I use 3000k LEDs as a bathroom light, but I dont use them during the day. The warm (low blue content) light at night, helps keep me relaxed so I can go back to sleep easily.

I use 4500k-5000k LEDs as worklights, in the evening, and also during the day. The cooler light (more blue content) helps keep me awake and alert.

===

because of the complexity of dual white balance scenarios… iow warm white contrasted to cool white, I do not show photos set to warm white. I only show photos set to cool white. This most closely illustrates what an operator will see when they get a new light and turn it on during the day, which imo, is the most common scenario from which we make initial color and tint observations.

Tint is also affected by the two different white balance scenarios, because sunlight has a higher DUV than incandescent.

therefore, a light with green tint, may not look very noticeably green, during the day

otoh, that same light can look much greener when observed during the evening, when the brain is adapted to incandescent.

iow, sunlight is not only cooler than incan, it is also greener tinted.

you can see how people could report apparently different experiences of tint, depending whether they look at the light when their brain is daylight adapted, as opposed to when it is incandescent adapted.

Yes!


I took a deep dive into photography in the early 2000s, completing a two-year course at a community college. I probably spent somewhere between one and two thousand hours color-correcting/printing both analog and digital images. Along the way, I learned how to see color, shadow and light rather than objects/things. Truly, it was a journey into sight.

Most of the time, I like to print my pix without a color cast. That means somewhere around 5200K-5500K, which is a bit of a "cheat" compared to the 5600K that characterizes midday sunlight. The lower CCT warms things up just a bit, which can "improve" skin tones.

To that end, I look for four things in a flashlight used to illuminate the subject of a photograph. If any of these are absent, I can try to fix them up in post-production, but it's much easier to get things right in the camera.

  1. CCT near the 5200K-5500K I am targeting for my finished images.
  2. Neutral tint, preferably right on the BBL.
  3. High CRI so that colors are rich.
  4. Smooth even beam without a hot spot. In practice, that often means using a TIR or a zoomie set to flood.

My background in photography has thus led me to reject the 4000K (and warmer) emitters preferred by many members of BLF. Of course, there is no right and wrong. I just happen to prefer neutral CCT in the low 5000K range. Your mileage may vary.

An Easy Way To "See" High CRI

Shine your lights on some blonde-colored wood furniture or an oak door. High CRI should be an obvious improvement over low CRI.

Lots of eerie sounds in those woods at night

Several caveats to this. The sun is not set to 5000K throughout the day. Golden/morning hour could easily hit 3000K-3500K with later morning hours hitting 4000K and overcast days hitting 6000K or higher.