Ultimate LED Bulbs - Ultra High CRI - The Honorable Quest

iamlucky13 · November 1, 2019, 5:41pm

It looks like the test data from Indie Cinema Academy is not the high CRI bulb, but Hyperikon’s standard CRI version.

Are you looking specifically for 240V bulbs?

In the US, 120V is almost universally standard.

iamlucky13 · November 1, 2019, 5:44pm

No, but warmer CCT’s do require a stronger red component to match their black body reference. This is difficult to achieve, especially without a hit to efficiency.

So even though a 4000K might have more red in its spectrum than a 5000K bulb, it could have a larger gap to its reference.

fneuf · November 2, 2019, 3:27pm

Same old story everywhere I believe : the fun with standards is there are many.

B22 (Swan's bayonet mount) had an original strong grip on UK (and Commonwealth) and France. But France always used both B22 (and a bit of B15) and E27 (and a bit of E14).

Now B22 mostly remain in usage in UK, India and Australia. France is "converting" (don't know of norms fixing that decision, seem to me it's a market general decision) to E27 since some years (decades), but you will still find some bayonet in old houses/rooms. The majority of Europe uses E27.

I don't have the full history of that, so it looks like a fully purposed mess. For instance I don't know why E26 and E27 both exists same time.

From their sole purpose (electric contact) they are mechanically interchangeables. So maybe it could have to deal with the 110 / 220v standard, because I think E26 is mostly in use in countries which adores 110v, whereas for E27 its 220v believers.Or could it be another cryptic example of metric vs imperial units (where 26mm would be the closest translation of some imperial measure). But this is pure speculation.

This could explain the really poor result at 17. Hyperikon's reference are not clear everywhere.

Well, well, well. Standards. Here we go again :)

Helpfully I have a map for that specific question:

Easy isn't it ? Well, we can still dig in and find more fun. More detailled map:

To each his own, standard, I guess :)

One of the great point of LEDs vs Incandescent or Halogens is that many manufacturers have cleverly took the opportunitu to build worldwide converters (90-250ish) in their bulbs. But not everyone of them.

Very interesting! I've also been recently presented an interesting article (sorry, it is in french) around the design choices of LED manufacturers to achiever High CRI in their chips

contactcr · November 1, 2019, 8:19pm

I had been trying to find a “smart” bulb that was also CRI 90+ only a month ago with not much luck

djozz · November 1, 2019, 9:21pm

Just measured the Yuji bulb: CCT=4040, duv=–0.0014, CRI=96.2, R9=91.6
So it does live up to its specifications.

One of my Philips GU10 3000K ExpertColor bulbs: CCT=2982, duv=+0.0004, CRI=97.4, R9=87.8

One of my Philips GU10 2700K ExpertColor bulbs: CCT=2779, duv=+0.0009, CRI=97.7, R9=88.5

fneuf · November 1, 2019, 9:49pm

Thanks for the work djozz. All your candidates have very great Duv values, special mention to the negative value of the Yuji.

Do you have R12 and Rf measurements also?

djozz · November 1, 2019, 9:57pm

I never used TM30 on my meter before, but I just added it to the parameters, had to measure again and had to look up what Rf and Rg were :innocent: :

Yuji 4000K: CCT=4035, duv=–0.0008, CRI=96.1, R9=90.9, R12=73, Rf=91.3, Rg=100.7

Philips ExpertColor 3000K: CCT=3009, duv=+0.0013, CRI=97.3, R9=89, R12=83, Rf=93.9, Rg=100.4

Philips ExpertColor 2700K: CCT=2796, duv=+0.0011, CRI=97.8, R9=89.6, R12=84, Rf=92.7, Rg=98.8

SKV89 · November 2, 2019, 6:54am

Waveformlighting 2400K filament light bulb
CCT: 2342K
DUV: –0.0032
CRI (Ra): 96.0
R9: 81.9
Rf: 93
Rg: 102
Blue Peak: 0.23

I absolutely love the negative DUV so it doesn’t look overly yellow. It is great as a nighttime light to minimize melatonin suppression. It is one of the lowest blue emitting lights I’ve measured, which is about right for the color temp but it still contains about 3x as much blue light as the E21A 2000K 9080.

fneuf · November 2, 2019, 10:51am

That might be a tad harder at the moment.

Last time I checked, top of the class was LIFX with close to 90 values (~86).

I have updated the first post table with your results, thanks again.

That's a shame really that there's no E26/E27 equivalent to those GU10 spots in Philips' current product range. They are not perfect, but very decent (R9 and R12 are still too low).

Thanks for your input SKV89, I have added your results to the first post table. The 0,23 blue peak does mean the overall curve has a nice shape and no, peak per-se, in the blue area?

lightdecay · November 2, 2019, 11:06am

Thanks for the test! I wanted to get a below-2700K LED bulb with negative duv, this one looks great.

fneuf · November 2, 2019, 3:17pm

Guys, if I want to create a "BLF Bulb Quality factor™" to "scientifically" order my first post table, what would you advise as metrics/formula? Which Rx, or other parameters to factor in?

Draft proposal => Qf = [(Ra *1)+(R9*3)+(R12*3)+(Rf*3)] / 10

djozz · November 2, 2019, 12:07pm

There is a point at which perfection can only be measured and not be seen at all by eye, are you sure you did not pass that point already?

patmurris · November 2, 2019, 3:03pm

I guess you mean France is converting from B22 to E27, and the majority of Europe uses E27?

fneuf · November 2, 2019, 10:14pm

That's a valid point, but from a number POV we have better offers from alternate manufacturers regarding R9 and R12 than the ExpertColors. Now, could I, could you, could the readers here, feel the differences, and what color rendering would they prefer? That's a good question.

To be sure about that, we'd need a setup with every "bulb contestant", and be able to blindly switch between the lights, and slowly build a prefered order. Kind of an ABX test adapted to lights infact.

But, isn't the title of this thread "Ultra High CRI"? And well, the subtitle of this thread could also be "Never accept compromises, to the infiniiiite and beyond..."

Slip of the tongue, thanks to point it to me! I'm correcting that.

Joshk · November 3, 2019, 6:17pm

I was actually thinking about this a few weeks ago. I think you are including the right things. But the best results would best be a simple scale from 1 to 10, but not try to fit the scale to the whole market today. For example, a 5 would actually be an above average light today, a 10 would be a future perfect light, and a 1 and 2 would be many of the lights on the market today. The idea of fitting the scale like that would so it doesn’t become obsolete by not accounting for improved technology in the next few years.

SKV89 · November 2, 2019, 5:58pm

yes the blue zone is very small, with blue peak is mild at 23% only.

djozz · November 2, 2019, 6:22pm

Talking about the blue peak, Luminus has brought a new 3030 led on the market that has the cyan gap filled up (or even made into a cyan peak) so that high CCT’s can be made with a lower blue peak. It is called Luminus “Salud”.