The Fabled Nichia 219B - still relevant?

Flashlight Modding and DIY Parts
49

What’s most surprising to me about the above pictures is the difference between the #2 and #3 beams. The LEDs are from the same tape and very closely shaved, both are TIR, but clearly the type of TIR and focus is causing pretty large variance.

50

I pm’d you my address, so I can take some photos in the format Im accustomed to. I use folded printer paper, line up warmest on left, iphone auto white balance, during the day (daylight ambient):

Color temperature of the LED and white balance of the camera each have an impact on the image, I always include a cool white light in my photos, as a reference

.

thanks for your photo, I increased the saturation and enlarged the labels, did not alter the tint

51

I was having a conversation with djozz and he filled me in about that… how the nature of a TIR can cause a visible shift in perceived temperature. And clearly in @contactcr’s beam shots, shaved emitters are clearly showing this too (the LED dome is an optic, after all).

It seems that a good many people attest to the Nichia 219B sw45k r9080 as a reference emitter for best CCT and CRI. I’m wondering — if you take a more efficient and powerful emitter with a color correction filter applied, would it still out perform (brightness, efficiency) and thus be preferable? The Samsung LH351D sure seems like a contender, especially with the right amount of dome shaving.

Was the 219B sw45k r9080 a ‘luck of the draw’ in LED production? That, akin to a fine wine, had the right lucky combination of materials selection and “conditions” in manufacture to achieve that result? And that intentionally trying to repeat it is like trying to repeat an excellent wine vintage — totally up to serendipity? Or… with manufacturers becoming more aware of the LED tint demands in commercial and enthusiast venues, that they’ll be able to find a way to recreate such a prized LED CCT + CRI?

52

Nah.

The SST-20 would be the best candidate if it were available dedomed from the start, as it is currently the only emitter(beside the E21A and Optilosis) to have R9080+ levels of color rendition.

If it was available in 5000k 95CRI variant, I think it would change a lot of stuff, especially available in good tint bins.

53

This is a super interesting thread to me.

Does this mean I have a problem? Haha.

Thanks everyone for contributing.

54

lol, I resemble that remark
for my entertainment, here is a supersaturated shot

tint above and below the BBL, thanks to maukka, pic is a link to his post

55

I always see people talking about the SS-T 20 Green tint and that's why I don't have anything yet with that led..

please tell me which lights comes with the SS-T 20 comes with the best tint available ( am interested in 3000k- 4000k no more)???

If I had to choose only one I choose the 4000k but I want flashlight with sst20 4000K with no green tint AT All

don't forget AM 100% nichia guy so please recommend me only the best please..

56

Thanks contractcr, to my eyes those two look identical. Very nice comparison, I have also seen TIR’s that shift to some completely odd ball tint. That’s pretty amazing how different TIR’s can make such a difference. It also amazes me how much the LH351D 5k changes to such a nice tint when shaved.
In jon_slider’s saturation pic of the “Mostly shaved LH351D, SMO reflector” it looks like it has a bit of green. I feel the closer its shaved also makes a difference in tint, I’m no expert but I see no green from shaving close with my samples in a SMO reflector.
Thanks for all the pics gentlemen, really educational.
I do feel that if more people tried the shaved LH351D it would become even more popular.
If I ever spend the time to figure out how to take tint comparison without looking completely different than what I see, I’ll post those up.
.
Djozz test with a shaved LH351C if anyone is interested. GB 351D closed. - #713 by djozz

57

word of the day
SuperSaturatedCaliFragiListicExpiAlyDosCious

58

I learned a new word today :smiley: .
So jon_slider when you are saturating the pics what exactly is that doing because I see nothing in person like in that pic above.
His pic looks close to what I see, the 219b looks pure white and the 351D looks slightly warmer both with a slight rosy tint. Just trying to figure out what the saturation pic does when I see nothing like that here.

59

With all the hype over the 219B sw45k it’s still way too cool for me personally. If there was an emitter that rosy at 3000-4000K I’d be all over it, but for now SST-20 and shaved LH351D do it for me for TIR and reflector lights respectively.

60

So in my picture all those tints are within probably ~600K CCT of each other (3975K~4600K -ISH)

The “mostly shaved” LH351D only looks green in comparison to the much lower DUV ones. If I only use that light by itself I cant notice any green at all. Also, the 219B SMO does look rosy cause it’s driven harder in single LED. The FW3A would have to be cranked up before it gets the “too rosy” (IMO) look.

  1. should be slight above BBL (but >= 4500K)
  2. RIGHT on BBL (like nearly 0.0 duv, and right around 3975-4000K)
  3. slightly below BBL (but = 4490K)
61

I’m still trying to figure out what a saturated pic shows, I have done some reading and I’m still not sure what that does.

From what I have read it depends a lot on what color temperature the original image was shot at, then there’s a whole can of worms you can open up to adjust color by saturation. The digital camera only knows the value of a color of RGB and any one has a value between 0 to 255 that all three values make up a certain color and how the camera sets the value of the image determines alot on what saturation can do to the colors. I’m not a photographer though and this stuff is way over my head.
:person_facepalming:

62

I think saturation just shows whatever is there, more intensely

it does not change the tint, and it does not change the color temperature
its just a way to turn up the colors, to make them more obvious, and exaggerated, imo

I agree it is very difficult to get a photo to look like what our eyes see, especially photos without a consistent white balance reference… I propose daylight white as the best reference, 5600k

what our eyes see, changes at night, compared to during the day, because of our brain using different white balance references

that is why I try to always include a cool white reference in my photos, and why I include a familiar Nichia…

I look forward to getting some first hand experience with the LH351, once I receive the light generously offered by contactcr. So far, I do see some pink in some of the pics posted…

I respect those of you with LH351 experience are getting more lumens than with a 219b, and that you seem happy with the CRI and tint.

63

Sometimes it feels like the world will end once the absolute perfect tint is finally found.

Prisoner - Fall Out revolt - YouTube, may the search always continue. :wink:

64

So shaving a 351 lowers the color temp, but what does it do to the CRI? Has anyone measured that?

65

I linked Djozz’s test with a 351C back in” post #56”:The Fabled Nichia 219B - still relevant? - #56 by moderator007.

66

Thanks a bunch!!

67

Instead of saturating images to show tint, I find it’s more revealing to normalize the brightness of each individual pixel. This basically eliminates differences in brightness, showing only the pure tint. The algorithm is really simple… just scale the brightness up until one of its color channels hits 100%. The color is otherwise unchanged.

Also, sometimes I blur the image first before putting it through the disco machine, to blend nearby pixels and reduce the impact of jpeg artifacts.

For example, given this beam shot someone posted on reddit… (I forget who; it seems to be deleted now)

Here’s what the disco machine spits out:

I think it’s a fine example of the “Cree rainbow” effect.

68

I see green and purple… I dont have your skill set to do pixel corrections… that looks like an XP-G3? Here it is saturated:

.

Thanks for sending me the Tool. The LED is nice, and by itself does seem slightly pink. Nice beam, and makes the palm of my hand look natural.

Only when compared to my sw45 and sw45k, does the LH351 look golden, not green though:

here is a saturated version

and here is a very rough attempt to correlate the saturated colors to a tint and color temperature chart

I think the LH351D, shaved or not, looks like a nice LED, thanks @contactcr for the generous sharing

Is the 219b still relevant? To me yes.