If mentioning it gets the message delivered of a modest tint shift from ok at low current to very nice at high current, then yes.
But there is āgreenā in the description, and ātint shiftā, so what sticks is different from what is intended. In that case it maybe more fair to the led not mentioning it at all
The XP-G3 4000K 90CRI and the XP-L2 4000K 90CRI are ugly green at all currents (as well as all 4000K 90CRI Osram Oslon leds newer than the very first generation, and the LG H35F0 4000K 90CRI), compared to which the LH351D 4000K 90CRI at low current has a wonderful tint. In fact several BLF-members are less than a fan of the rosiness of the Nichiaās, for them this Samsung led may have the perfect balance in the tint.
And perhaps it needs mentioning once more: in my experience, once using a flashlight in the real world instead of your wall, CRI has clearly more impact than tint, the tint must be really off to make a flashlight unpleasant to use.
I found some tint shift measurements by maukka, and the LH351D doesnāt appear to be any worse than other emitters. So, probably not worth mentioning.
Tracking says a prototype should arrive in the next 18 hours, so the plan is to do some testing on that and then put the emitter poll online. I hope the comparison data is in order now, but thereās still a little time to make corrections. Also, Iām not sure if the flood/throw pic should be included or not.
Does it have to boil down to just one LED flavor? or is it possible to have two versions right from the start?
What I gather from the discussion of the last few days is that a single LED flavor will almost inevitably leave a good (perhaps even large) portion of potential buyers unhappy.
With two version its much more likely to please the majority.
My suggestions for that would be these two:
XP-L HI 5000K (highest throw, highest lumens, but low CRI & high cost)
Samsung LH351D 4000K 90CRI (high CRI, no extra cost, but least throw) [if the Samsungs are unobtainable for some reason > Nichia 219C 4000K 90CRI]
I rarely use my torches in daylight, but mostly at night, or to peer into dark places.
And I prefer light levels that donāt upset my dark-adapted eyes too much, rarely at turbo levels, often at firefly or moonlight level.
Meaning my eyes are probably operating in the mesopic or even scotopic range, rather than photopic.
I wonder whether a lot of tint preference is under artificial conditions, e.g. bouncing off white walls at close range and at high levels, whereas under my sort of usage at much lower illumination levels my eyes see the tint very differently.
I much prefer the warmer tints, 4000K or below, with good CRI, for me the difference is very marked, I donāt like cool white at-all, neutral or warm works much better. My theory is that the warmer tints, more biased towards red rather than blue, compensate for the Purkinje shift towards blue sensitivity at lower illumination levels, keeping the colours more ānaturalā looking.
Edit: which is why it is so difficult to take realistic photos of outdoor night-time scenes, the digital camera does not respond the same way as the human eye. Perhaps someone could develop a Purkinje shift plugin for photoshop to simulate the effect with a set of curves for different wavelengths.
I realise that those who prefer to light things up āas bright as dayā may have different preferences.
I donāt have any experience with the LH351D 4000K 90CRI, but I value Djozzās opinion. If he likes it, Iām sure it would be a great choice. Perhaps it would be best to keep the light as cheap as possible, and not go with the xpl-hi, at first. Iām sure the cheaper it is in the beginning, the more successful it will be. If its a huge hit like Iām certain it will be, hopefully more flavors will be released later with different emitter choices. This formula has worked will for the Emisar line.
Seconded. 4000K. The floodier beam is a big plus for me too, when out searching for things that could be anywhere, possibly brown-coloured, or dripping red stuff.
This is the result of a relatively broken voting method called plurality voting. It does not capture more complex real-world preferences like āI prefer 5000K, but I wouldnāt be upset with 4500 or 4000, both of which I strongly prefer to 3500, but even that is better than 6500ā.
There are voting systems that do capture such preferences, and TK will be using one for the poll.