I canāt compare them right now, but for what itās worth, here are the ID codes from the Phargos Ebay store:
Oslon Square 2700K = LCW CQAR.CC-MQ-6U-L2
Oslon Square 3000K = LCW CQAR.CC-MP-6R-L1
Oslon Square 3500K = LCW CQAR.CC-MP-6Q-MX
Oslon Square 4000K = LCW CQAR.CC-MR-6L-L2
These are all āCRI 96ā versions and theyāre all beneath the BBL, which is what most of us would prefer.
Unfortunately they donāt have 4500K in CRI96, nor in CRI95, nor 92, nor 90ā¦
Avoid dramatizing with this fellows, remember that conjetures or assumptions based quarreling is usually a nice waste of time.
This stuff is not as simple, imho. You may think emitter tints slightly above or below the black body locus may not look exactly white and this is true, but that is a coordinate parameter probably obtained from some sort of average output deviation at the different measured frequencies. A high CRI emitter is going to have a much closer spectral distribution emission curve to that of a blackbody or sunlight, I believe this means a ābadā high-CRI tint just means the light will have a subtle tint, probably even more subtle than a lower CRI emitter with a closer to the black body locus tint. 
And please remember you're not just passive ideal observers but creators of your own bubbles of reality. 
Take care 
A high CRI emitter can have a significantly off white tint, like this green Oslon which is one of the worst (duv 0.0093) Iāve tested regardless of CRI.
However, as the tint was quite consistent throughout the beam in the test light (Thrunite TC12 V2), it can be corrected with a simple filter (unlike an XP-G3 for example).
White means that a tint is the on the BBL. As long as the intensity of the light is matched to the CCT (see here) the human brain will get used to it and perceive such light as āwhiteā.
.
These 8T and 8Q tints are worse than what Clemenceās Nichia LEDs do (see Maukkas tests). The current Nichia 219C and 144A R9050 LEDs with the ecxeption of the 4000K (3-step 403) variant are all rather good. Many people here seem to prefer an off-white reddish/pinkish tint which seems to make colors āpopā more. The 219B-V1 sw45 and the 219C 4070e are the only LEDs here until now which garantuee such a tint.
.
I also prefer such off-white tints, but only inside. Outside I only like them when my surroundings are very brownish (forest with no leaves etc.). Otherwise a bit more balanced light is more pleasing.
.
The 144A 4500K R9050 in particular has a very natural tint, very close to the BBL. Unfortunately is suffers from the same problems as current Cree LEDs (the corona has a lower CCT than the hotspot and the spill).
Iām always totally puzzled by Maukkaās graphsā¦ :person_facepalming:
You should get help 
( 
).
Here is his thread on the topic. It makes sense to discuss details there.
First you need to decide what you want to know when looking at his measurements. Then post those questions and you will get help in finding the answers in his measurements.
Me too, but they sure are neat looking.
maukka, it may as well be possible that the way we perceive light is the actual reason for our very fine kermit discrimination tendencies. Just look how intense just about any cheap green laser looks (they may hit 0.1W) versus vastly more powerful blue ones.
On green! Nice shot! :-D
Cheers ^:)
I agree with this. Even the green lantern like Osram can be made pleasant with simple gel filter. What Iāve learnt from Leeās LED filter is you can always trim but you canāt add. High CRI LED is a better starting point than lower one when you plan to use filters.
Maukka, do you have Lee filter swatch? Iām curious how much can we trim the blue spike in most cool white LED while keeping the rest as flat as possible.
- Clemence
I like Lee minus green filters
I donāt understand duv, yet, but, I can see that the Greens (R3, R4, R5 and R11) are taller than the Red, Yellow, and Pink (R9, R10, R13), on the Green tinted Oslonā¦ while on the Pink tinted Nichia the Red, Yellow and Pink bars are not shorter than the Green barsā¦
images clipped from maukkaās plots
looking at the relative size of the red vs green bars, I would expect all these LEDs to have a green tint
and they do: (pic is from, and is a link, to maukkaās review)
I havenāt dig too much into this butā¦.IMHOā¦.
Simply said Duv is how far off the tint of tested LED from the BBL (0 - 4500K) or SBL (5000 - 10000K). BTW, is there such a term as Sun Body locus? 
this is just my own word.
From my understanding looking at Maukkaās chart: 0,0055 means the tint is lower than BBL while 0,0093 means the tint is higher than BBL/SBL. By carefully looking at his charts, I think itās safe to conclude anything beyond 0,004 or 0,004 is considered too far from the reference line (BBL/SBL). This aligns very well to 219B sw45k (miss piggy tint) and Oslon (kermitt tint) facts.
- Clemence
I dont know how I missed this ledā¦ but I didā¦ is it still available anywhere to purchase?
The reference source differs depending on the standard/method. For example the CRI standard CIE 13.3: 1995 describes the duv tolerance (i.e. acceptable duv to consider the CRI valid in the first place) as +ā0.0054 with BBL being the reference source for CCTs up to 5000K and D series illuminants above it.
This is also true for CRI, CQS and CRI2012, even if the duv tolarence varies a bit.
In the TM-30-15 and CIE 224:2017 the transition from BBL to daylight illuminants is gradual.
In my data, however, duv is always the distance from the BBL, even on sources above 5000K.
I have the Lee swatch book to play with.
Weāll be very grateful to see your test results then. My personal request is how far can we trim the blue spike. Also how the minus blue spike filter affects CRI.
Many thanks
Clemence
Mauri, the CRI datas are missing
Where do you get those green filters? I have several LEDās that could use that.
Lee offered a free small swatch colour sample, it consists of hundreds of its colour gel filter. Each sample is enough for two small flashlights. Just contact your local representative marketing. I lost my colour swatch in the project site. Iāll need to order another
- Clemence
Who is Lee?
