Suggestion on developing high CRI LED flashlight

[degarb]

No not really. Cameras can do just fine with <100 CRI, even with fairly poor CRI, though somewhat flat spectrum helps. Biggest issues are monitor calibration, trade-offs in the RGB pattern demosaicing, differences between capture filters and display spectrums, etc. Trying to catch the subtlety of color with only three broadband sensors results in trade-offs, though in theory if the capture and display have the same spectral characteristics, you can get very close. You are adding in narrow band reflectance issues of the objects you are viewing/etc.

As opposed to high-cri, dual white point LEDs systems are being used which creates more data points.

Most people are happier with colors that are more saturated than reality and would pick those almost always anyway when it comes to photography.

Dual white. You mean like my crude trick of wearing a cool on head and warm of wrist, for color skip/streak hunting while painting? Or visa versa.

Can you think of any useful situation where the Yuji on battery would be worth the premium?

My concern too is reduced luminance per watt will be a regression in color perception, not an improvement. Unless, runtime does not matter and the Yuji could be pumped furiously with current.

Sorry to say, just got a new Nichia 219b 4000 Kelvin flashlight in the mail yesterday. Compared to my xpl hi, the only situation it was better was upclose on red and red heavy wood (even compared to the warm xpl hi) . At any distance greater than 10 foot, it was crushed by the 4300 Kelvin xpl hi for color, as it was just plain dimmer. All my research keeps saying, color is cri, gai, and luminance, combined. Both the Nichia and xpl hi 4300 Kelvin light, shift the powder blues toward teal, when they are pure blue. And I got to hand it, the 6500 xpl hi is way better than most xml from the 2014 Era and before, still selling inside cheap lights.

All tests were done at 2.1 amps, New Panasonic 3400s. I will test current with alternate mm later. Hoping my early impressions are correct, and not way off what is achievable with the hi cri emitter. If the Nichia could be pumped double current from my tests, it might keep up with the other two, and surpass in color rendering for the eye.

Really, the high cri Nichia has brought me no closer to the perfect task light. Still need that cool for good blue. And I am not sure about the swap choice of lumens for better red.

 

[degarb]

Testing current next morning with a good multimeter, the Nichia may have been driven at 75% the xpl hi. Probably 70 percent as efficient, these 4300kelvin go to v5 bin. This would put the Nichia at 475 lumen v. 900 lumen cool xpl and probably 800 is for the warm. The xpl could have a more efficient driver, who knows. This would account for the non-trivial difference between the two emitter brands. So, still an agnostic. I would still be compelled, against my will, to buy a Yuji light, just for professional curiosity. That is, unless specs looked horrible.

 

[ssanasisredna]

Dual white. You mean like my crude trick of wearing a cool on head and warm of wrist, for color skip/streak hunting while painting? Or visa versa.

Can you think of any useful situation where the Yuji on battery would be worth the premium?

My concern too is reduced luminance per watt will be a regression in color perception, not an improvement. Unless, runtime does not matter and the Yuji could be pumped furiously with current.

Sorry to say, just got a new Nichia 219b 4000 Kelvin flashlight in the mail yesterday. Compared to my xpl hi, the only situation it was better was upclose on red and red heavy wood (even compared to the warm xpl hi) . At any distance greater than 10 foot, it was crushed by the 4300 Kelvin xpl hi for color, as it was just plain dimmer. All my research keeps saying, color is cri, gai, and luminance, combined. Both the Nichia and xpl hi 4300 Kelvin light, shift the powder blues toward teal, when they are pure blue. And I got to hand it, the 6500 xpl hi is way better than most xml from the 2014 Era and before, still selling inside cheap lights.

All tests were done at 2.1 amps, New Panasonic 3400s. I will test current with alternate mm later. Hoping my early impressions are correct, and not way off what is achievable with the hi cri emitter. If the Nichia could be pumped double current from my tests, it might keep up with the other two, and surpass in color rendering for the eye.

Really, the high cri Nichia has brought me no closer to the perfect task light. Still need that cool for good blue. And I am not sure about the swap choice of lumens for better red.

Everything you are saying here makes complete sense. Color discrimination is better at higher lighting levels which makes high CRI at low lighting levels a curiosity but not a necessity. I could add in that at low levels, we tend to prefer warmer colors, which further throws your discrimination of blues out the window if followed.

I expect Yuji will just get an existing flashlight supplier to put their emitter in it and slap their name on it with perhaps a few UI tweaks, so development cost is likely to be low. There are likely people who will get pulled in by the high-cri whether there is a good objective reason to or not to cover the development. That said, serious lighting customers will wonder "why" they would go down this avenue unless it is purely for giveaways.

 

[KITROBASKIN]

Guessing that distributors are not so interested in these Yuji LED's, thinking that demand is not strong enough, and so Yuji is making an effort to make their product known through us.

 

[KITROBASKIN]

Let's hear an update from Daniel.

 

[neutralwhite]

Got this reply from Yuji;

In fact, we are recently facing some problems in producing flash lights.

Our LEDs do not have lens on the surface.

Let's hear an update from Daniel.

 

[KITROBASKIN]

Thank You for providing that information!

(neutralwhite: your mailbox is full)

 

[neutralwhite]

Cleared thanks.

Thank You for providing that information!

(neutralwhite: your mailbox is full)

 

[Keitho]

People obviously have opinions all over the map. If Yuji wants the largest possible appeal on one light, make it easy for opinionated flashaholics to change on their own: optics, battery tube, switches, and driver programming. If Yuji sells their high CRI LED on a thick multi-LED board with a good open source driver and multiple options for battery sizes, inside of a decent host, there will be lots of sales to the general public, modders, and collectors. The host should have more heat sinking than the D4 to allow for some massive heat off of the high-CRI LED. A solid copper head might be really cool-looking and effective...

 

[iamlucky13]

Got this reply from Yuji;

In fact, we are recently facing some problems in producing flash lights.

Our LEDs do not have lens on the surface.

Curious. When I look at their datasheet for the 3030 series, the spatial distribution graph looks almost identical to an XP-L HI, as I suspected it might. I actually quickly overlayed the two in an image editor, and once scaled the same, the lines overlapped for a lot of the graph.

Granted, the emitting surface is smaller than that of an XP-L HI, but I kind of expected it would still work fairly well in a small reflector.

 

[ssanasisredna]

People obviously have opinions all over the map. If Yuji wants the largest possible appeal on one light, make it easy for opinionated flashaholics to change on their own: optics, battery tube, switches, and driver programming. If Yuji sells their high CRI LED on a thick multi-LED board with a good open source driver and multiple options for battery sizes, inside of a decent host, there will be lots of sales to the general public, modders, and collectors. The host should have more heat sinking than the D4 to allow for some massive heat off of the high-CRI LED. A solid copper head might be really cool-looking and effective...

Is "lots" more or less than 1000? General public for the most part wants lumens, and 90CRI is more than good enough, and more lumens is often better. Modders and collectors? ... that is not going to even get a business started let alone sustain one.

 

[ssanasisredna]

Curious. When I look at their datasheet for the 3030 series, the spatial distribution graph looks almost identical to an XP-L HI, as I suspected it might. I actually quickly overlayed the two in an image editor, and once scaled the same, the lines overlapped for a lot of the graph.

Granted, the emitting surface is smaller than that of an XP-L HI, but I kind of expected it would still work fairly well in a small reflector.

Could be a near field / far field issue. Far field they are the same, near field, where an optic would come into play, they could be different.

 

[neutralwhite]

Any update from Yuji now ?.