Nichia NV4L144/W144 AME/ARE CRI and output testing

Clemence sent me a bunch of new NV4 L144/W144 AME/ARE Nichia LEDs for color testing.

Here’s the list of all the emitters tested here:
NV4L144AME 6V 3000K R8000
NV4L144ARE 12V 3000K R9050
NV4W144AME 6V 5700K R9050
NV4W144AME 6V 6500K R70
NV4W144ARE 12V 5700K R9050
NVSL219CT 3000K R9050

More specific bins are listed in the spectrum graphs.

All the 6V AME emitters were tested by per Nichia’s specifications at 6V / 1400mA while the 12V ARE emitters at 12V / 700mA. The NVSL219CT was tested at 3V / 700mA. Djozz will test the maximum output later with higher currents.

Without any reflector, the color temperature is way higher than specced when measured right in front of the LED. That’s why I used the orange peel reflector from my Nitecore EC4SW (MT-G2). Its size isn’t optimal for the Nichias, but the only suitable thing I had in hand. For the 219CT I used a smaller reflector from an Astrolux S1. With the EC4SW reflector, the hotspot is clearly warmer than the surround corona and spill. This behavior was observed with all of the NV4 emitters as can be seen from the graphs below.

This is why the reflector or optics has a big role. Comparison between tint with a reflector (hotspot), a bare emitter straight on and at 45/90 degree angles and the LED totally diffused.

For color rendering I used a diffuser which nicely averages the total output and gives the best approximation.

Even though not as visible as in the XHP-series from Cree, the Nichias do have distinct four diodes.

On to the results:
Overall comparison between tints using a diffuser.

Nichia NVSL219CT 3000K R9050

Tint in different parts of the beam using a reflector from an Astrolux S1.

Tint diffused, with reflector and without from three different angles (0° straight on)

Nichia NV4W144ARE 5700K R9050 12V

Nichia NV4W144AME 5700K R9050 6V

Nichia NV4W144AME 3000K R8000 6V

Nichia NV4W144ARE 3000K R9050 12V

Nichia NV4W144AME 5700K R9050 6V

Nichia NV4W144AME 6500K R7000 6V

Output and forward voltage test

Thank you for your review. Without this information no LED should be allowed to come out on the market.

Hi Mauri,

Great thanks for the test =) This is very valuable for all of us.
Sorry for my late reply. Just got home from a long motor bike trip today. As expected, the color tests showed very nice result. I don’t have a tool to validate it but my real life eye test also confirmed that too.
Yes, the high angle tint collected by the reflector or TIR (in my test) project warmer tint in the beam spot too. The best optics would be those designed for colour mixing. My all round favourite is the 5700K 9050.


Thank You maukka!

I added the CRI to your overall diffused tints

and sorted your list by CRI
NVSL219CT 3000K R9050 = CRI 94
NV4L144ARE 12V 3000K R9050 = CRI 93
NV4W144AME 6V 5700K R9050 = CRI 91
NV4W144ARE 12V 5700K R9050 = CRI 91
NV4L144AME 6V 3000K R8000 = CRI 84
NV4W144AME 6V 6500K R70 = CRI 75

Very nice, these look hopeful.

Any idea on lumen output?

@ Jon Slider: Great! now the chart looks even more useful

The 3000k R9050 is perfect for Incan replacement. What’s more: less eye strain due to less blues too

Soon, after Maukka send them to Djozz =)

Excellent information, thanks.

Can’t wait for the 219CT R9050 to become available.

Thanks for the thorough testing maukka.

I will try my best to do a proper voltage and light output test on a few of these, but unfortunately the timing is very bad: next week monday to friday I will be supervising a school trip to London, and the week thereafter monday to friday I’m away with the family on a short holiday. I will try to squeeze out one output test next weekend, but no promises. Serious hobby time is two and a half weeks from now :frowning:

Impressive tests!



219C and 144A


People do not give nearly enough credit to high CRI lights. Ever since trying them I can’t go back. They are just SOOOO much nicer to look at and even with less lumens you can see much better with them.

Now my big thing is tint mixing, tint mixing high CRI LED’s takes it up another level.

Actually mixing super LOW CRI LED is more fun :smiley:
You’ll never know what you’ll get. Grey could be magenta, yellow turns green, etc…
It’s like the witch and her kettle

Now there’s an interesting idea.

Has anyone done tests like the ones maukka did in this thread, but with a tint-mixed light? Perhaps one each of 5700K, 5000K, 4500K & 3000K, all 90+ CRI, on a quad MCPCB and diffused together into the CRI testing equipment?

I have not seen any testing like this but would love to. Some 90+ CRI 3000k + ~5700k 90+ CRI would be quite interesting.

To my eyes it is richer then a plain ~4500k 90+ and has more contrast but I don’t really have any directly comparable setups.

I do not see how a tint mixed light consisting of high CRI leds of various colour temperatures would emit a noticably different (better, closer to 100CRI) spectrum than one high CRI led in the averaged tint. Of course it is fun to make a mixed light (my S41 has two 5000K and two 3000K 219C’s), but I would not expect something special new, the resulting spectrum is a simple add-up of the separate spectra of the leds.
As maukka mentioned, mixing several low CRI leds can have an interesting and possibly high CRI result, but if you already use high CRI leds, the CRI difference between separate leds and the resulting mixed spectrum will become much lower. No miracles are happening in light spectra.

Not miracles by any means but to my eyes it is more contrasty then a single tint high CRI LED.

That said I don’t have a multi-emitter high CRI light in a single tint to compare so it is hard to say for sure.

It is kind of like HDR imaging, you are able to artificially increase the contrast and vibrancy above that which is normally possible.

But then all of this is why I would love to see some real world testing on it to see what the actual differences are.

My kitchen lamp, which I inexpensively retrofitted with two of these Markerled 10W warm white 5730 lamp boards, plus a single “18-36W” driver (300mA, 18-36 led/ledpacks in series) driving a couple of 30S2P cool white 5730 arrays in parallel (so ≈150mA of current flowing through each array, 75mA/emitter), actually displays quite a nice tint, warmer than daylight (phone's pics look slightly warmer when taken under the lamp with “daylight” white balance), and CRI is, to my naked eye (and my friend's one too) noticeably better than the one from a cool white XM-L2.

My ¢2.

Cheers ^:)

Yup, I think so too. CRI can’t be combined like CCT or lumen. There has to be proper mix/rato between each spectrum’s for a particular CCT and CRI. My thinking goes like this, For example: simply add a 6500K cool white and 2700K super warm “normal” LED would result in approximately 4600K with worse CRI than before because now the 480nm spectrum region curve would be even deeper in contrast to those to the left and right scale.

Most high CRI normal LED’s spectral curves are very different from the true high CRI LEDs like Yuji, Ledengin, Nichia, Citizen (vivid), etc… They have fatter than average curve in the 480nm and yellow - red zone. Adding several LED like these would result in less CRI rating decrease compared to the skinny curved high CRI LED.

Note: This is just my thinking, not really sure until someone publish a valid test.

Put the CRI aside, the mixed results is indeed interesting and perceived differently depends on each set of individual eyes :partying_face:


thank you for those excellent beamshots! They are perfect for demonstrating what high-cri really means because you used two cool white LEDs and didn’t move the camera at all. It’s a very obvious difference.

I think tint-mixing can have some effect if one combines warm white and cool white high-cri LEDs. Cool whites have more blue and more cyan (most LEDs produce very little cyan) and warm-white LEDs produce more red light. It would make a lot of sense to use a warm-white LED that produces a lot deep red light (the other area of the spectrum that most LEDs are not good at). Doing this should result in a more balanced spectrum, but with too much green and yellow (both LEDs would emit a lot of this).

Another way of doing things might be to use a triple optic and combine one high powered cool-white high-cri LED with one cyan and one deep red led (with less power). This would basically make the spectrum more even without adding more yellow and green light.