Why don't my 2 Nichia 291B flashlights look anything alike?

I have two flashlights, both supposedly built with the same LED (4500K High-CRI Nichia 219B). One is a Surefire 6P with a Nailbender/Customlites drop-in and the other is a Surefire 9P with the Sportac triple drop in. I've had the 6P/Nailbender drop in for a couple of weeks now and the Sportac drop in just showed up today. In going from one to the other to play with them, I noticed that the light that comes out of them looks NOTHING alike! To provide background to that statement, I worked as a professional photographer and professional photographic printer for quite a long time, so I'm well used to looking at light and seeing what's really there rather than what my brain wants to assume is there. Additionally, I have accidentally found a fantastic test of LED color rendering. The board game Takenoko has 3 colors of tiles. Low-CRI (meaning less than 85 or so) light sources make it almost impossible to tell the pink tiles from the yellow tiles. High-CRI (meaning over 90) will not only show which is pink and which is yellow, they'll reveal some of the color variation within the tile.

So, the 6P/Nailbender is very cold/blue, much colder than I was expecting for a 4500K LED. In fact, it's slightly colder (visually) than the 5000K CREE A19 light bulb I have. It barely passed my quick-n-dirty color rendering test in that you could identify the pink and yellow tiles but it's wasn't as easy as it should be and you couldn't see any variation in the color on the tile.

The 9P/Sportac, on the other hand, looks, well, normal for what it claims to be. It pretty well matches what I'd expect to see out of a 4100-4500K source with a high CRI. The light reads as white, not blue-tinted and it passes the color rendering test with flying colors. Literally, if you throw the tiles!


I'm used to the idea that LEDs are binned into pretty coarse groupings and so some variation between examples is to be expected. However, I have never seen two LEDs claiming to be the same thing that are this staggeringly far apart.

Is it possible one of them is a mixup and had the wrong LED installed on the dropin?

I don't think the 219B LEDs are HCRI 90+ when tested.

The Sportac drop-in uses Nichia 219A, whereas the Nailbender is a 219B. They are different LEDs, although similar. Both claim to be high CRI but the 219A is warmer and gives better rendition of long-wavelength colours.

RedForest UK said:

The Sportac drop-in uses Nichia 219A, whereas the Nailbender is a 219B. They are different LEDs, although similar. Both claim to be high CRI but the 219A is warmer and gives better rendition of long-wavelength colours.

Click to expand...

The last data from Eagletac said they'd switched to the 219B for the Sportac dropins this year.

williaty said:

Additionally, I have accidentally found a fantastic test of LED color rendering. The board game Takenoko has 3 colors of tiles. Low-CRI (meaning less than 85 or so) light sources make it almost impossible to tell the pink tiles from the yellow tiles. High-CRI (meaning over 90) will not only show which is pink and which is yellow, they'll reveal some of the color variation within the tile.

Click to expand...

That probably has as much to do with the colour temperature of the light source, as it does with the CRI. Warmer tints will show reds (pinks) and yellows better, at the same CRI or perhaps even a worse CRI.

As mentioned, the Nichia 219A is warmer than the 219B, and is likely a better LED for showing colour in most situations.

Is it possible you received old stock? I have a Nichia 219 in an older M31 Malkoff that I like a lot. It is better than the Nichia 219 I had in an M61, but not nearly as different as you describe. That seems like quite a variance.

WalkIntoTheLight said:

That probably has as much to do with the colour temperature of the light source, as it does with the CRI. Warmer tints will show reds (pinks) and yellows better, at the same CRI or perhaps even a worse CRI.

As mentioned, the Nichia 219A is warmer than the 219B, and is likely a better LED for showing colour in most situations.

Click to expand...

Again, Sportac has switched to the 219B early in 2014. So it's a B-vs-B comparison here.

One thing I did just notice is that Nailbender/Customlites makes this dropin with two different varieties of 219B. One is a 4500K/92CRI and the other is 5000K/90CRI. I ordered the 4500K version. The dropin was mis-labeled as a multi-mode (I ordered single mode and that's what it turned out to be). That makes me wonder if this is actually the 5000K version that I received even though it was supposed to be the 4500K version.

If the 6P/NB one was the 5000K version, at the cold edge and low-CRI edge of the allowable bin and the 9P/Sportac one was the 4500K version at the warm edge and high-CRI edge of the allowable bin, that might explain the very large apparent difference.

williaty said:

Again, Sportac has switched to the 219B early in 2014. So it's a B-vs-B comparison here.

One thing I did just notice is that Nailbender/Customlites makes this dropin with two different varieties of 219B. One is a 4500K/92CRI and the other is 5000K/90CRI. I ordered the 4500K version. The dropin was mis-labeled as a multi-mode (I ordered single mode and that's what it turned out to be). That makes me wonder if this is actually the 5000K version that I received even though it was supposed to be the 4500K version.

If the 6P/NB one was the 5000K version, at the cold edge and low-CRI edge of the allowable bin and the 9P/Sportac one was the 4500K version at the warm edge and high-CRI edge of the allowable bin, that might explain the very large apparent difference.

Click to expand...

According to the Nichia datasheets, it is possible to find some 219b with cold tint. In this case, the CRI would be a bit lower.

Consider too the reflector may have an effect on color temp. If you've ever used umbrellas to bounce flash then you're familiar with the effect a gold lining has vs a silver lining on the umbrella.
I was comparing a mule light (white surface) against a reflector light (silver surface) and wondering if the reflecting surface didn't account for the difference I was seeing (another case of same LED/different rendering).
Just a thought.