CREE's neutrals have an ugly tint. Why all the fanfare for XPG or XML neutral lights?

[Esko]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

I have close to 1,000+ lumens with my Kerberos quad Nichia 219 drop in. "Considerably reduced output" isn't even a consideration here.

A very nice light. However, a similar quad XP-G2 configuration would give you ~2000 lumens (and less heat), so, the considerably reduced output is still there.

Nice photograhps. What is the white balance setting in them?

Good color rendering helps in color perception. However, more power helps, too. Lower illumination means less saturation in colors (especially in mesopic vision).

Seeing color rendered through "pure" neutral light (sunlight 5500 Kelvin) is a hardwired component of our brain.

...

Reminder: Warm and cool wavelengths are not "NATURAL"

The sunlight can vary from some 2000K to more than 10000K depending on conditions.

 

[AnAppleSnail]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML

It's well and good to be a tint, cri, and color perception snob. But show me such a flashlight with equally effective output, light distribution, runtime, and size. We'll each find our favorite compromise along the dimensions of beam profile, output, runtime, light quality, and size.

But there is no single best CCT or CRI. Study after study shows situations where various tasks are best served by various CCT. The effects of improved color perception are usually found to be marginal outside color matching. Show me a bear-fleeing person who can't tell leaves from roots by an adequate Cree tint bin (and CRI>85) and I'll show you someone who won't tell a bear from a barn by looking

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

A very nice light. However, a similar quad XP-G2 configuration would give you ~2000 lumens (and less heat), so, the considerably reduced output is still there.

Nichia 219 performs better than you (and I) thought.

This chart puts my 5.6A quad Kerberos at about 1,300LED lumens using 1.4A per LED

Nice photographs. What is the white balance setting in them?

Whatvever setting is on the Sony RX100. The BEST slimline digital camera on the market. Features a full 1" CMOS sensor. What it photographed, is what I see.

Good color rendering helps in color perception. However, more power helps, too. Lower illumination means less saturation in colors (especially in mesopic vision).



The sunlight can vary from some 2000K to more than 10000K depending on conditions.

Ideally mid-day temperature is 5500K and a good benchmark to measure artificial "pure" lighting against.

 

[MichaelW]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML

Who said that?

Philips Lumileds, Cree.
Cameras can mislead, so you have to try it yourself.

 

[jorn]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

Whatvever setting is on the Sony RX100. The BEST slimline digital camera on the market. Features a full 1" CMOS sensor. What it photographed, is what I see.

lol. If it's the best camera, you should be able to turn off auto settings like iso, shutterspeed, wb. That way the camera wont change the settings from one pic to the next. If not, dont compare.

I can make any of the pictures "look best" if i adjust some on my tv/monitor. It's not like a printed picture. Monitors are just like the tint lottery. Some are good some are crappy, and a lot are crappy adjusted. How do i know that i see what you see, when we look at the same picture but on different monitors?

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

lol. If it's the best camera, you should be able to turn off auto settings like iso, shutterspeed, wb. That way the camera wont change the settings from one pic to the next. If not, dont compare.

I can make any of the pictures "look best" if i adjust some on my tv/monitor. It's not like a printed picture. Monitors are just like the tint lottery. Some are good some are crappy, and a lot are crappy adjusted. How do i know that i see what you see, when we look at the same picture but on different monitors?

What part of "what it photographed is what I see" did you not understand?

Conversely you can call every photo (or video) posted on candlepowerforums into question with your line of reasoning.

 

[jorn]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

What part of "what it photographed is what I see" did you not understand?

Conversely you can call every photo (or video) posted on candlepowerforums into question with your line of reasoning.

I understand cameras. It is for a reason pepole post the camera settings with the pictures, or state that the camera was locked with manual settings when they got several pictures to compare.
I just dont belive your "what it photographed is what I see" claim to be 100% true when it comes out of a point and shoot auto mode. Put all the lights in one picture if you cant lock the camera. Auto mode might be like comparing different flashlights with different tinted sunglasses on..
Thats why you should not rely too mutch on point and shoot pics when comparing lights. You got incredibly good eyes and memory if you can go outside take a pic, walk in, and and say for sure: "Thats exactly what i saw 5 min ago." Put the monitor issue on top of it, and yeah. Pics are not a great tool for a true tint snob.. Even with a SLR and photoshop, it's really hard to the get the tints close to 100% on your own monitor.

 

[Esko]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

Nichia 219 performs better than you (and I) thought.

This chart puts my 5.6A quad Kerberos at about 1,300LED lumens using 1.4A per LED

You forgot to mention the source of your table, which is here (thanks to foxtrot824). A similarly powered (5,1 watts per led) XP-G2 flashlight would emit ~2150 led lumens (1,6A, ~3,15V, 537 lumens/led).

Whatvever setting is on the Sony RX100. The BEST slimline digital camera on the market. Features a full 1" CMOS sensor. What it photographed, is what I see.

It is a nice camera, but it looks like it's automatic white balance is poor. I took some test photos with my 7 years old Canon 350d in December. With automatic white balance, the results were quite similar with XP-G 5000K, Luxeon Rebel high cri 5000K, XM-L 4500K and Nichia 219 4500K high cri. There were some minor differences in color rendering, but white was white in every photograph and colors looked pretty much right, too. There was a barely visible green hue in photographs taken with XM-L cool white and XP-G S3 cool white (which is famous for it's green tint) and the photo of Luxeon Rebel high cri 4200K looked a bit yellow. The only photographs with a clearly biased white balance were XP-G 3000K high cri (yellow) and unknown 5mm cool white (angry blue). I must add that with corrected white balance, the high cri XP-G rendered colors very well, too.

Of course, with fixed white balance the photographs would have clear tints. What kind of tints depends heavily on the value that is chosen.

Ideally mid-day temperature is 5500K and a good benchmark to measure artificial "pure" lighting against.

Nichia is 4500K, and many cool whites are 6500K.

Anyway, the illuminance of the mid-day sunlight is tens of thousands of lux. The illuminance of a flashlight is usually very different.

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

You forgot to mention the source of your table, which is here (thanks to foxtrot824). A similarly powered (5,1 watts per led) XP-G2 flashlight would emit ~2150 led lumens (1,6A, ~3,15V, 537 lumens/led).



It is a nice camera, but it looks like it's automatic white balance is poor. I took some test photos with my 7 years old Canon 350d in December. With automatic white balance, the results were quite similar with XP-G 5000K, Luxeon Rebel high cri 5000K, XM-L 4500K and Nichia 219 4500K high cri. There were some minor differences in color rendering, but white was white in every photograph and colors looked pretty much right, too. There was a barely visible green hue in photographs taken with XM-L cool white and XP-G S3 cool white (which is famous for it's green tint) and the photo of Luxeon Rebel high cri 4200K looked a bit yellow. The only photographs with a clearly biased white balance were XP-G 3000K high cri (yellow) and unknown 5mm cool white (angry blue). I must add that with corrected white balance, the high cri XP-G rendered colors very well, too.

Of course, with fixed white balance the photographs would have clear tints. What kind of tints depends heavily on the value that is chosen.



Nichia is 4500K, and many cool whites are 6500K.

Anyway, the illuminance of the mid-day sunlight is tens of thousands of lux. The illuminance of a flashlight is usually very different.

I should be clear, the ONLY photos posted in this thread using the Sony RX100 are the six in this post:
http://www.candlepowerforums.com/vb...utral-lights&p=4130517&viewfull=1#post4130517

Everything else from the strawberry's to the bush and field shots are from various places on CPF and BudgetLightForum. God only knows what awful cameras they where using for those.

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML

The holy grail of LED flashlight BIN and tints, IMO

Can anybody figure out what it is?

For anyone wondering the light source is the "SUN"

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

You forgot to mention the source of your table, which is here (thanks to foxtrot824). A similarly powered (5,1 watts per led) XP-G2 flashlight would emit ~2150 led lumens (1,6A, ~3,15V, 537 lumens/led).

Troof! thanks for that. Was too lazy to find the location of that chart from awhile ago.

There's also no argument the XP-G2 outperforms the Nichia 219 in luminance. The 219 still outperforms the XP-G series in CRI. It's still a milestone LED in its own right. And 4 of them loaded into the head of a Surefire C2 fueled by an unprotected 3400mAh Panasonic makes for a "pocket torch" like no other.

 

[Esko]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

I should be clear, the ONLY photos posted in this thread using the Sony RX100 are the six in this post:
http://www.candlepowerforums.com/vb...utral-lights&p=4130517&viewfull=1#post4130517

Everything else from the strawberry's to the bush and field shots are from various places on CPF and BudgetLightForum. God only knows what awful cameras they where using for those.

Yes, you should indeed be more clear and tell people when you are posting photographs that are not your own. Or use links. Give the credit to those who deserve it.

Just out of curiosity, I changed the white balance in my test photographs to 5500K. Nothing looks very nice any more. The best of the ugly group seems to be Luxeon Rebel 5000K high cri with XM-L 4500K following as a close second. Everything else looks pretty horrible (Nichia 219 high cri too). What surprised me though was that when I changed it to 4500K, XM-L 4500K looked actually good. Rebel 5000K was pretty good, too. Everything else had ugly hues. While I guess Nichia was the third in both cases (with a clear difference to the first two), I think this is a nice little reminder that camera settings can make a difference.

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

Yes, you should indeed be more clear and tell people when you are posting photographs that are not your own. Or use links. Give the credit to those who deserve it.

Just out of curiosity, I changed the white balance in my test photographs to 5500K. Nothing looks very nice any more. The best of the ugly group seems to be Luxeon Rebel 5000K high cri with XM-L 4500K following as a close second. Everything else looks pretty horrible (Nichia 219 high cri too). What surprised me though was that when I changed it to 4500K, XM-L 4500K looked actually good. Rebel 5000K was pretty good, too. Everything else had ugly hues. While I guess Nichia was the third in both cases (with a clear difference to the first two), I think this is a nice little reminder that camera settings can make a difference.

Agreed. This is where personal testimony from individual contributors can chime in and add to a growing body of evidence that points in a particular direction. I've made my contribution.

 

[ShineOnYouCrazyDiamond]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

Ya know what I think the problem with the tint lottery is? I goes back to what I learned in college in my EE classes. Resistors are binned with 1%, 5% and 10% tolerances within the stated spec Ohm rating. If you want a resistor with 1% tolerance you are going to pay for it. The way they do it is that out of a batch all the resistors within 1% get moved to that bin, everything 1% to 5% go to the 5% bin and everything 6-10% goes to the 10% bin. So you know going in that if you cheap out and buy the 10% bin you will never get anything better than 5%.

I think you can apply the same logic to LEDs. LEDs are tint binned when they come off the line. Ideal tints that run along the ANSI ideal white curve are harder to come by and cost a significant premium when purchased in bulk. LEDs that are outside this curve are in the majority and get binned in wider tint variation ranges further off the ANSI curve. These are sold in bulk to mass manufacturers at a heavy discount and put in production flashlights.

Long and short of the argument is you get what you pay for.

Take a look at the ANSI White curve w/ Cree subdivisions here: https://www.candlepowerforums.com/threads/156772 and you will clearly see the S, T, R and U bins on the outer edges of the curves. These are the cheap LEDs that manufacturers buy for less. Check out the pricing on the various XPG tints here http://www.cutter.com.au/products.php?cat=Cree+XPG and you can see that there is a premium for the more desirable tints.

At the end of the day I consider myself an extreme neutral tin snob but I never rub anyone else's face in it or tell them what tint they should like. For me my tint is the Cree 5C1/5B4. I will take this tint any day all day long and love it.

YMMV :ironic:

 

[GunnarGG]

Re: CREE's neutrals have an ugly tint.

Ya know what I think the problem with the tint lottery is? I goes back to what I learned in college in my EE classes. Resistors are binned with 1%, 5% and 10% tolerances within the stated spec Ohm rating. If you want a resistor with 1% tolerance you are going to pay for it. The way they do it is that out of a batch all the resistors within 1% get moved to that bin, everything 1% to 5% go to the 5% bin and everything 6-10% goes to the 10% bin. So you know going in that if you cheap out and buy the 10% bin you will never get anything better than 5%.

I think you can apply the same logic to LEDs. LEDs are tint binned when they come off the line. Ideal tints that run along the ANSI ideal white curve are harder to come by and cost a significant premium when purchased in bulk. LEDs that are outside this curve are in the majority and get binned in wider tint variation ranges further off the ANSI curve. These are sold in bulk to mass manufacturers at a heavy discount and put in production flashlights.

Long and short of the argument is you get what you pay for.

Take a look at the ANSI White curve w/ Cree subdivisions here: https://www.candlepowerforums.com/threads/156772 and you will clearly see the S, T, R and U bins on the outer edges of the curves. These are the cheap LEDs that manufacturers buy for less. Check out the pricing on the various XPG tints here http://www.cutter.com.au/products.php?cat=Cree+XPG and you can see that there is a premium for the more desirable tints.

At the end of the day I consider myself an extreme neutral tin snob but I never rub anyone else's face in it or tell them what tint they should like. For me my tint is the Cree 5C1/5B4. I will take this tint any day all day long and love it.

YMMV :ironic:

I think you're right about tolerances and prices.

Does anybody know how big the price difference actually is if you choose only the best tints with narrow tolerances?

If I buy a light for let's say $100 with a large risk of loosing the tint lotteri, how much extra would I have to pay to be sure to get the expected tint (narrow tolerances)?
I don't know how much it is but I guess it's worth it.

I rather pay $110 and have a light that I like with a nice tint than pay $100 and have the same light but with a ugly tint.

I'm not talking about sending the light away for modding, I'm talking about the manufacturer putting a little more effort and money into getting good tints.

 

[ShineOnYouCrazyDiamond]

Re: CREE's neutrals have an ugly tint.

Maybe somebody with industry experience could speak better to the pricing point.

I only buy batches of 5-10 LEDs cut off a reel so I definitely pay a premium. Manufacturers buy reels of 1,000s.

 

[jtr1962]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

Ya know what I think the problem with the tint lottery is? I goes back to what I learned in college in my EE classes. Resistors are binned with 1%, 5% and 10% tolerances within the stated spec Ohm rating. If you want a resistor with 1% tolerance you are going to pay for it. The way they do it is that out of a batch all the resistors within 1% get moved to that bin, everything 1% to 5% go to the 5% bin and everything 6-10% goes to the 10% bin. So you know going in that if you cheap out and buy the 10% bin you will never get anything better than 5%.

There's actually a bit more to it than this. The reason higher tolerance resistors cost more is because they need to be made more stable. A resistor which might be within 1% of nominal at room temperature does nobody any good if goes 5% either way at temperature extremes. Because of this, high tolerance resistors are usually manufactured separately. There generally isn't any issue of what to do with resistors which aren't within 1% of nominal because standard 1% values are much closer to each other than standard 5% values (see chart). For each tolerance category every resistor made will be able to be sorted into one of the standard values. I've noticed over the last few years that 1% resistors aren't any more expensive than 5% resistors. Even 0.1% resistors aren't completely off the charts like they used to be. You can get certain values for under 10 cents in 1000s.

LEDs are a bit different. The manufacturer always aims for the highest flux bin and the most idea tint (for a given color temperature). Due to process variations not all LEDs make the cut, so they're sorted into bins. I would imagine some percentage fail to make any bin. Those are either discarded or perhaps sold to manufacturers using them in low-cost items where tint really isn't all that important (i.e. flashlights and lanterns sold a dollar stores).

 

[jtr1962]

Re: CREE's neutrals have an ugly tint.

Does anybody know how big the price difference actually is if you choose only the best tints with narrow tolerances?

It's not as large as you might think. If you look at the various part numbers for Cree LEDs sold by Mouser, decipher them for which are tighter tolerances, you'll find maybe a 10% or 20% price difference. The price differences are actually much larger between flux bins, with the highest flux bins always fetching a big premium because of their relatively scarcity. I think LED manufacturers have actually gotten pretty good at making consistent tints, so they no longer fetch a huge premium. Certainly the days of getting very blue or purple or green LEDs are over because those wouldn't even make the cut in the loosest bins.

 

[ledmitter_nli]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

There's actually a bit more to it than this. The reason higher tolerance resistors cost more is because they need to be made more stable. A resistor which might be within 1% of nominal at room temperature does nobody any good if goes 5% either way at temperature extremes. Because of this, high tolerance resistors are usually manufactured separately. There generally isn't any issue of what to do with resistors which aren't within 1% of nominal because standard 1% values are much closer to each other than standard 5% values (see chart). For each tolerance category every resistor made will be able to be sorted into one of the standard values. I've noticed over the last few years that 1% resistors aren't any more expensive than 5% resistors. Even 0.1% resistors aren't completely off the charts like they used to be. You can get certain values for under 10 cents in 1000s.

LEDs are a bit different. The manufacturer always aims for the highest flux bin and the most idea tint (for a given color temperature). Due to process variations not all LEDs make the cut, so they're sorted into bins. I would imagine some percentage fail to make any bin. Those are either discarded or perhaps sold to manufacturers using them in low-cost items where tint really isn't all that important (i.e. flashlights and lanterns sold a dollar stores).

Another frequent Main Street, StarBucks Denizen? Small world!

And here too so it seems
http://www.candlepowerforums.com/vb...u-re-wearing&p=4132736&viewfull=1#post4132736

 

[jtr1962]

Re: CREE's neutrals have an ugly hue. Why all the fanfare for XPG or XML neutral ligh

Another frequent Main Street, StarBucks Denizen? Small world!

Actually, I'm more into the Chinese food when I go to Main Street. Great food, nearly as good as Chinatown.

And here too so it seems
http://www.candlepowerforums.com/vb...u-re-wearing&p=4132736&viewfull=1#post4132736

Yeah, we could start our own Flushing branch of CPF here! Small world indeed!