High CRI leds, comparing R9 values, XP-G3 N219c N218b XM-L2 XP-L2

90 posts / 0 new
Last post

Pages

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia
High CRI leds, comparing R9 values, XP-G3 N219c N218b XM-L2 XP-L2

collected images from maukka, showing that 219b beats 219c and XP-G3 and XM-L2 and XP-L2, in red rendering. Focus on the height of the Red Bar (R9)

And a couple Zebras.. you can see they are very low in R9, which imo is the reason there is so much “tint lottery”. Its not really lottery, its just that the High CRI Zebras have very low R9.. Not a lot of red, so the green and yellow dominate their tint.

References:
links to maukkas original spectrum charts that are the source of the images I cut and pasted.
Any errors, omissions, and misinterpretations of the original images, are my own:
Olight S1 Mini Baton CW+HCRI review with measurements (XM-L2/XP-G3, 16340) | BudgetLightForum.com
Lumintop Tool AAA Copper + Titanium measurements (Nichia 219B) | BudgetLightForum.com
Zebralight SC5Fc measurements (AA, warm tint) | BudgetLightForum.com
219c 4000k 9050 chart by maukka from Clemence LEDs at Virence
219b 4500k 9080 chart by maukka from Clemence LEDs at Virence

Edited by: jon_slider on 04/04/2018 - 18:04
AnhTran
Offline
Last seen: 1 month 2 weeks ago
Joined: 05/13/2017 - 23:56
Posts: 108
Location: Australia

I have a Convoy s2 with triple Nichia 4500k R9090, colour tint is amazing

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia

Very nice! I have the same LED in my pocket
In the middle, on right is sw45 9080 Smile

kiriba-ru
kiriba-ru's picture
Offline
Last seen: 55 min 18 sec ago
Joined: 01/17/2016 - 02:34
Posts: 1702
Location: Russia

1. 400Lm led vs 800-1200Lm leds
2. 5 years old led vs modern leds
3. Looking only at R9 is not good idea. Even one point at 2-axis tint graph have more correlation with real feelings.
4. You are not right, this is tint lottery. Each cree led is different, and this range is too big. I wont wonder if Zebralight orders reels with “tint package” (named like E2-E5).
5. Zebralight work with cree many years. Better color rendering is not enough to leave this business relations and choose another vendor that may be worth in all other aspects

Need a vlogger for machine accessories reviews

Barkuti
Barkuti's picture
Offline
Last seen: 49 min 16 sec ago
Joined: 02/19/2014 - 14:46
Posts: 2776
Location: Alhama de Murcia, Spain

Interesting to see how saturated blue (R12) “sucks” in all of them, too. Likely because of the “cyan dip”. 

 

Cheers Smile

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia

based on the above comparisons of High CRI LEDs, my observations are:
Lights with lower R9 have a larger proportion of yellow and green than red.

The N219b has more R9 than newer LEDs such as the XP-G3 and N219c.

Brighter LEDs have less R9 and more R10 and R11.

The XP-G3 and 219c are brighter and greener than the 219b.

Zebras have very low R9 and are know for having yellow and green tint.

Zebras are brighter and greener than N219b lights.

The N219c is brighter and greener than N219b.

The_Driver
The_Driver's picture
Offline
Last seen: 10 hours 6 min ago
Joined: 10/20/2016 - 05:51
Posts: 1125
Location: Germany

A R9-value of 40-60 is not bad, it’s much, much better than 95% of standard LED flashlights (most have a negative R9-value). Cree LEDs have a tint problem, not CRI.

Of course the R9080 LEDs from Nichia look better, but part of the reason is their better tint. They are either on the BBL or below it. Tint is much more strongly perceived than small parts of the spectrum.

The best ones are probably the Nichia E21As with R9080. They don’t have a dome which improves angular tint consistency.

Osram also has “R9080” LEDs (declared as CRI95), but the ones with 4000K and above have a yellow tint. So they look much worse than some of the Nichias.

Barkuti
Barkuti's picture
Offline
Last seen: 49 min 16 sec ago
Joined: 02/19/2014 - 14:46
Posts: 2776
Location: Alhama de Murcia, Spain

The_Driver wrote:
A R9-value of 40-60 is not bad, it's much, much better than 95% of standard LED flashlights (most have a negative R9-value). …

Heard this before but, how is a negative value possible? Suffice to say I do not understand how the CRI colour scores work, of course, but to me a 0 is a 0 which would mean no emission in these frequencies. 

 

Cheers Smile

The_Driver
The_Driver's picture
Offline
Last seen: 10 hours 6 min ago
Joined: 10/20/2016 - 05:51
Posts: 1125
Location: Germany
Barkuti wrote:

The_Driver wrote:
A R9-value of 40-60 is not bad, it’s much, much better than 95% of standard LED flashlights (most have a negative R9-value). …

Heard this before but, how is a negative value possible? Suffice to say I do not understand how the CRI colour scores work, of course, but to me a 0 is a 0 which would mean no emission in these frequencies. 


 


Cheers Smile

It becomes obvious when you check out here how the individual Ri values are actually calculated (specifically Step 9).

mrheosuper
Offline
Last seen: 2 hours 36 min ago
Joined: 09/30/2016 - 12:44
Posts: 1191
Location: Ho Chi Minh City, Vietnam

now i understand why i love 219b so much, even more than 219c

Forgot my pen

staticx57
Offline
Last seen: 2 hours 21 min ago
Joined: 04/11/2016 - 00:43
Posts: 464
Location: New Jersey, United States

SBT70 5700k HiCRI

219C R9050 5700k

The_Driver
The_Driver's picture
Offline
Last seen: 10 hours 6 min ago
Joined: 10/20/2016 - 05:51
Posts: 1125
Location: Germany

Thank you!
Finally someone has measured the SBT-70 high-cri!

I own a light with this LED. It’s as good as the R9080 Nichias, but of course way brighter, but also way more inefficient.

Luminus states an R9 of 85 for the square version of this LED (CBT-90).

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia

The_Driver wrote:
Cree LEDs have a tint problem, not CRI.

Thank you, I agree Tint, CRI, and Color Temperature are all independent variables.
Also agree that Low CRI LEDs have Negative R9, which may help explain why they make red things look brown, and the palm of my hand look lifeless.

Specific to the High CRI charts I posted, it is coincidentally also true that the lights with the lowest R9 happen to also have the Greenest tint (above the BBL), and the lights with the highest R9 happen to also have the most Rosy tint (below the BBL).

Here is a Low CRI light, courtesy of maukka, with Negative R9, that happens NOT to have Green tint, in fact it lands perfectly on the BBL.

But this is a Low CRI LED, that makes Red things look Brown, and I did not make this thread to talk about them. The point of this thread was specifically to compare the R9 values of some common and popular High CRI LEDs.

I did not expect to discover that the High CRI LEDs Zebralight uses happen to have some of the lowest R9 of the group, but of course Zebras are focused on Brightness, not CRI. Most of the flashlight market is brightness driven, and Low CRI is common in the majority of offerings. So is Green tint. It seems to me that the N219b is one of the few offerings that tends to avoid Green tint, but not always.

Im such a snob about tint, that I find N219b is a personal favorite, particularly the 9080 versions.

staticx57 wrote:
SBT70 5700k HiCRI… 219C R9050 5700k

thanks for adding to my knowledge. Will SBT70 work in an S Mini (w XM-L2 Tir)?:-)
And can you share a link to where you found those charts please?

Here is another group of R9 comparisons, for some of the 219c 2700k, 3000k, 3500k, qnd 4000k offerings from Clemence

These all happen to have more yellow/green tint than the N219b I posted earlier, and coincidentally the 219c also have R9 bars that are shorter than the two 219b in the first post.

The_Driver
The_Driver's picture
Offline
Last seen: 10 hours 6 min ago
Joined: 10/20/2016 - 05:51
Posts: 1125
Location: Germany

Staticx57 probably measured that himself. I have never seen a measurement other than the one in the datasheet (it includes a simple spectrum, looks the same as here).

This LED is not produced anymore though and was hideously expensive (50$ when not on sale) and inefficient.

NeutralFan
NeutralFan's picture
Offline
Last seen: 1 day 4 hours ago
Joined: 03/20/2014 - 19:22
Posts: 671
Location: Wisconsin, USA

So what’s the best high CRI, high R9 LED that you can buy today? I hate to ask since it may tempt me to build another flashlight, but if there’s another LED that’s significantly better than the Nichia 219C D240 on 20mm MTN DTP CU – 90+ CRI 4000K, I’m certainly interested.

I’d rather use my flashlight around the house than turn on the lights.

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia

People in this group buy thread bought up the grail N219b LEDs such as the sw45 9080, maybe ask for help there

this chart originally by maukka shows and example set of 219C that he tested.

They are available here
https://www.virence.com/product-page/nichia-nvsl219ct-nvsw219ct

Tjohn
Tjohn's picture
Offline
Last seen: 42 min 40 sec ago
Joined: 12/22/2015 - 07:22
Posts: 255
Location: mouth of the Shark rift
NeutralFan wrote:
So what’s the best high CRI, high R9 LED that you can buy today? I hate to ask since it may tempt me to build another flashlight, but if there’s another LED that’s significantly better than the Nichia 219C D240 on 20mm MTN DTP CU – 90+ CRI 4000K, I’m certainly interested.

Unless you run across any 219C Deviants, your best bet, as Jon stated, would be to go with 219b 9080 (sw40 and sw45) LEDs from the group buy, since they have a pedigree.
-
Currently available:
Maukka showed the 219c sm303 9050s from Clemence as coming closer to the ANSI BBL than the 4k and 3.5k
(sm403 and sm353).
You can always get lucky with 5 and 7 step Nichias when hunting for R9, as Jon can attest.
Lower luminance rated leds (D200 and D220) are normally a better choice for red/rosey in the 4k and under range.
As mentioned, it boils down to personal taste, with many insisting the lower CRI 219c sm505 with higher luminous flux appear to have more red in their tint than the sm503 9050, therefore YMMV…….Remembering a reflector can make a great difference, along with the level an LED is driven.

The_Driver
The_Driver's picture
Offline
Last seen: 10 hours 6 min ago
Joined: 10/20/2016 - 05:51
Posts: 1125
Location: Germany

A reddish tint does not mean that the R9-value is high!

Tjohn
Tjohn's picture
Offline
Last seen: 42 min 40 sec ago
Joined: 12/22/2015 - 07:22
Posts: 255
Location: mouth of the Shark rift

The_Driver wrote:
A reddish tint does not mean that the R9-value is high!

Certainly not for a technophile….however, do you have a better ‘experiential’ manner of elucidation for a layman?
Tjohn
Tjohn's picture
Offline
Last seen: 42 min 40 sec ago
Joined: 12/22/2015 - 07:22
Posts: 255
Location: mouth of the Shark rift

The_Driver wrote:
Staticx57 probably measured that himself. I have never seen a measurement other than the one in the datasheet (it includes a simple spectrum, looks the same as here).

This LED is not produced anymore though and was hideously expensive (50$ when not on sale) and inefficient.


Interesting presumption to say the least, regarding somebody offering information?
Tjohn
Tjohn's picture
Offline
Last seen: 42 min 40 sec ago
Joined: 12/22/2015 - 07:22
Posts: 255
Location: mouth of the Shark rift

The_Driver wrote:
A R9-value of 40-60 is not bad, it’s much, much better than 95% of standard LED flashlights (most have a negative R9-value). Cree LEDs have a tint problem, not CRI.

Of course the R9080 LEDs from Nichia look better, but part of the reason is their better tint. They are either on the BBL or below it. Tint is much more strongly perceived than small parts of the spectrum.

The best ones are probably the Nichia E21As with R9080. They don’t have a dome which improves angular tint consistency.

Osram also has “R9080” LEDs (declared as CRI95), but the ones with 4000K and above have a yellow tint. So they look much worse than some of the Nichias.


Does CREE have a problem beyond tint, venturing into credibility?
-
Does apologizing for CREEn make 40-60% R9 less objectionable?
-
How does that E21a look in a most existing reflector lights?
Might be less than forceful unless you gang them up, with resultant crosses and other aberrations?
How does this LED fit into the needs and intention of the OP, and person wanting to know what is available for his particular project?
Fifty cents each is a bargain to have them reflowed by Clemence, but he lives in Bali Indonésia.
Billy X
Billy X's picture
Offline
Last seen: 2 days 43 min ago
Joined: 02/03/2013 - 05:50
Posts: 556
Location: South of France

My E21’s arrived from Clemence today,ordered on 8th march,so no worse than typical chinese vendor,and more reliable.

The_Driver
The_Driver's picture
Offline
Last seen: 10 hours 6 min ago
Joined: 10/20/2016 - 05:51
Posts: 1125
Location: Germany
Tjohn wrote:
The_Driver wrote:
A reddish tint does not mean that the R9-value is high!
Certainly not for a technophile….however, do you have a better ‘experiential’ manner of elucidation for a layman?

For example you can have a flashlight with a perfect “snow” white tint, but when you light things up with this light, colors look “dull”. If specificallY red objects look dull, then the R9 value is probably low.

Tint is something you can see when you light up a white surface.

The r9 value can only be seen when you light up red (maybe also dark brown) colored objects. Basically how much the color “pops” (stands out).

Tjohn wrote:
The_Driver wrote:
Staticx57 probably measured that himself. I have never seen a measurement other than the one in the datasheet (it includes a simple spectrum, looks the same as here).

This LED is not produced anymore though and was hideously expensive (50$ when not on sale) and inefficient.


Interesting presumption to say the least, regarding somebody offering information?

That is a weird comment! When somebody posts a measurement like that it’s natural to presume that he made it himself (he didn’t state the source). Maukkas measurements look different (his software calculated all of the values).

Tjohn wrote:
The_Driver wrote:
A R9-value of 40-60 is not bad, it’s much, much better than 95% of standard LED flashlights (most have a negative R9-value). Cree LEDs have a tint problem, not CRI.

Of course the R9080 LEDs from Nichia look better, but part of the reason is their better tint. They are either on the BBL or below it. Tint is much more strongly perceived than small parts of the spectrum.

The best ones are probably the Nichia E21As with R9080. They don’t have a dome which improves angular tint consistency.

Osram also has “R9080” LEDs (declared as CRI95), but the ones with 4000K and above have a yellow tint. So they look much worse than some of the Nichias.


Does CREE have a problem beyond tint, venturing into credibility?
-
Does apologizing for CREEn make 40-60% R9 less objectionable?
-
How does that E21a look in a most existing reflector lights?
Might be less than forceful unless you gang them up, with resultant crosses and other aberrations?
How does this LED fit into the needs and intention of the OP, and person wanting to know what is available for his particular project?
Fifty cents each is a bargain to have them reflowed by Clemence, but he lives in Bali Indonésia.

I really don’t understand the intention of some of your comments. “Apologizing for Cree”?!? I stated simple facts, that is all. Most white LEDs have 65-75CRI and very low, usually negative R9-values. That is needed for very high efficiency. So Cree high-cri LEDs with an R9-values of 40-60 and CRI of >=90 are indeed much better than standard LEDs. That’s just obvious.

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia

Just to clear up any confusion, these are the two factors I wanted to investigate:

Which of the 10 charted LEDs has the highest R9?
Which of the 10 charted LEDs has the most Rosy tint?

Highest R9 in 219b, lower R9 in 219c and Cree.

Coincidental Observations:
The Higher the R9 the Rosier the Tint
The lower the R9 the Yellow/Greener the Tint.
the pattern is consistent for all 10 High CRI LEDs I was comparing.

staticx57
Offline
Last seen: 2 hours 21 min ago
Joined: 04/11/2016 - 00:43
Posts: 464
Location: New Jersey, United States

Tjohn wrote:
The_Driver wrote:
Staticx57 probably measured that himself. I have never seen a measurement other than the one in the datasheet (it includes a simple spectrum, looks the same as here).

This LED is not produced anymore though and was hideously expensive (50$ when not on sale) and inefficient.


Interesting presumption to say the least, regarding somebody offering information?

Yes I did measure these myself using the software CT&A and an XRite i1 Pro Spectrophotometer

Tjohn
Tjohn's picture
Offline
Last seen: 42 min 40 sec ago
Joined: 12/22/2015 - 07:22
Posts: 255
Location: mouth of the Shark rift
The_Driver wrote:
…….Of course the R9080 LEDs from Nichia look better, but part of the reason is their better tint. They are either on the BBL or below it.


Although the sw45 219b 9080s are shown by Nichia to be on or below the ANSI BBL….the lower temperature 219b 9080s straddle the BBL.
Most 3 step 219c Nichias are centered slightly above, using the more modern binning system.

The rose/pink/magenta phosphor mix of the sw45 9080s look unreal to some, and like eye candy to others.

NeutralFan
NeutralFan's picture
Offline
Last seen: 1 day 4 hours ago
Joined: 03/20/2014 - 19:22
Posts: 671
Location: Wisconsin, USA
Tjohn wrote:
Remembering a reflector can make a great difference, along with the level an LED is driven.

I also found the lens can make a big difference too. I built a Convoy S2+ with a Nichia 219C R9050 90+ CRI 4000K, and with the non-AR glass lens the beam was good, but there was an annoying yellow ring around the hotspot (ie. corona). I also noticed that there was a small crack that formed on the LED dome, which created an artifact in the beam.

So I decided to make a diffused lens from some acrylic plastic, but it made the tint a little more yellow.

I then sanded down one side of the glass lens to make it frosted and the resulting beam turned out awesome. No more corona, just a beautiful slight rosy tint. It’s more of a floody beam, but there are no artifacts at all.

I’d rather use my flashlight around the house than turn on the lights.

jon_slider
jon_slider's picture
Offline
Last seen: 4 hours 27 min ago
Joined: 09/08/2015 - 12:20
Posts: 1347
Location: Crowdifornia

The_Driver wrote:
…….Of course the R9080 LEDs from Nichia look better, but part of the reason is their better tint. They are either on the BBL or below it.

I agree. I prefer the tint of my N219b over the others.
4000k N219b 9050 Lumintop Tool:
maukka wrote:

and sw45 9080:
maukka wrote:

My 4000k N219b 9080, and 9050, are less pink than my 4500k N219b 9080

Tjohn wrote:
Although the sw45 219b 9080s are shown by Nichia to be on or below the ANSI BBL….the lower temperature 219b 9080s straddle the BBL.

I agree, the tint charts and my experience confirm that.

sw45 9080 from Clemence

Kitchen window overcast daylight 7pm in California, iPhone 8

sw 40 9050 from Illumn

kiriba-ru
kiriba-ru's picture
Offline
Last seen: 55 min 18 sec ago
Joined: 01/17/2016 - 02:34
Posts: 1702
Location: Russia

jon_slider wrote:
Just to clear up any confusion, these are the two factors I wanted to investigate:

Which of the 10 charted LEDs has the highest R9?
Which of the 10 charted LEDs has the most Rosy tint?

Highest R9 in 219b, lower R9 in 219c and Cree.

Coincidental Observations:
The Higher the R9 the Rosier the Tint
The lower the R9 the Yellow/Greener the Tint.
the pattern is consistent for all 10 High CRI LEDs I was comparing.


No.
90CRI XP-G3 have much bigger R9 rate than most 70-80 CRI cree leds. But they look much more yellow/greener than xp-g2 3d for example.

Need a vlogger for machine accessories reviews

staticx57
Offline
Last seen: 2 hours 21 min ago
Joined: 04/11/2016 - 00:43
Posts: 464
Location: New Jersey, United States

Jon,

Id like to offer you another data point. Here is an LED I swapped into a Preon P1 (Run on high). Oslon 3000k 95 CRI. To my eyes this has both too much yellow and a tiny bit too much green. This LED will fit anywhere you have a Nichia 219X

Also the only cool white LED id consider, the Nichia E21A 6500k R9080

Tjohn
Tjohn's picture
Offline
Last seen: 42 min 40 sec ago
Joined: 12/22/2015 - 07:22
Posts: 255
Location: mouth of the Shark rift

The first two images, at differing levels, show a (TIR) PT16 with 3 Nichia 219b sw45 9080 emitters from Clemence on the left vs. 6 Nichia 219b sw45 9050 (92+ CRI) in an Eagtac MX253LC reflector light.

The third photo (at lower output) has an Eagtac D25 AAA with the same small die Nichia 219b sw45 9050 emitter, this time using a TIR.
The differences in CRI, R9, and location relative to the BBL are major variables.
All photos were daylight balanced.

Pages