Cree T8 4ft LED Review

[cybrdyke]

Cool. I just proved that.......

Why post that?
It sounds like you're back to your conspiracies again....everyone is lying about all their data.

32 watt lamps dont deliver 2800 lumens. I must admit I'm disappointed that you dont know this basic tenet of fluorescent lighting.
2 T8 lamps in a typical fixture deliver somewhere around 3100 to 3600 lumens.
So your 50% light loss claim isn't correct.
CD

 

[Platonic Solid]

Why post that?
It sounds like you're back to your conspiracies again....everyone is lying about all their data.

32 watt lamps dont deliver 2800 lumens. I must admit I'm disappointed that you dont know this basic tenet of fluorescent lighting.
2 T8 lamps in a typical fixture deliver somewhere around 3100 to 3600 lumens.
So your 50% light loss claim isn't correct.
CD

Being condescending does nothing for your credibility. Exactly what are your qualifications in this area?
When comparing published lamp data of both types I am correct. To get into actual lumen output once any lamp is installed into any fixture will vary greatly based on fixture type, lamp placement, luminaire reflectivity, reflector shape, lens transmittance (if it has one), … in other words, stating that a typical 2 lamp fluorescent fixture delivers “X” lumens is a pointless statement without providing lumen output for an identical fixture with 2 LED lamps.

 

[Denwood]

Gents, the reason I did this little review was to provide "real world" results using the products in my shop. I can do a quick power test with one fixture using my EM100 comparing both bulb types. Just keep things friendly please

 

[cybrdyke]

Being condescending does nothing for your credibility. Exactly what are your qualifications in this area?
When comparing published lamp data of both types I am correct. To get into actual lumen output once any lamp is installed into any fixture will vary greatly based on fixture type, lamp placement, luminaire reflectivity, reflector shape, lens transmittance (if it has one), … in other words, stating that a typical 2 lamp fluorescent fixture delivers “X” lumens is a pointless statement without providing lumen output for an identical fixture with 2 LED lamps.

I'm not really interested in your opinion of my credibility. I am friendly until provoked. Twice you've tried. And I do find your claims of ulterior motives by competent manufacturers as Oliver Stone-ish. If you want to have a gentlemanly conversation, we can.

You are wrong, sir. True there can be difference between the performances of luminaires. But mainstream luminaires that use these types of lamps are all pretty similar. You CAN draw some conclusions without specific fixture info. It doesn't vary that much.
There is no instance where you will get 2800 lumens out of a T8 lamp with a standard ballast, because the ballast factor for standard light output is .88. Second because a typical troffer has a fixture efficiency anywhere from .65 to .8 . In a 2 lamp troffer, 2800 x 2 x .88 x .7 = 3449 delivered lumens.
The 1600 lumen output of the InstantFit lamp (not 1500 as you stated, probably just a typo), is WITH the .88 ballast. Since the beam angle on the LED tube is 160, which closely mimics the beam angle of a standard troffer, there is no fixture efficiency multiplier. In a 2 lamp troffer, 1600 x 2 = 3200 delivered lumens.
Any lensing issue is moot because both types of lamps will be behind the same lens.

 

[cybrdyke]

Gents, the reason I did this little review was to provide "real world" results using the products in my shop. I can do a quick power test with one fixture using my EM100 comparing both bulb types. Just keep things friendly please

Cool. I'm sure you'll find the numbers come out just as they should.
Happy New Year.
CD

 

[Platonic Solid]

Just curious:
Which brand do each of you have in actual usage?

http://jamesindustry.com/
ZY-T8-22W1200 4000K Frosted to be exact.
4ft lamps run about $10 each, but shipping is a killer unless you're buying bulk (5000+).
They're UL Classified and DLC Listed.
Avoid the clear lens as they really look nasty in any fixture = too much harsh point light source.
They also make ballast compatible lamps which have the benefit of being usable with or without a fluorescent ballast.
My future plan for the downstairs drop ceiling in my house is the flat panel edge lit. Not as energy efficient, but a truly beautiful, perfectly even light output.
I have several other brands running in test fixtures at work, plus I make my own (but those aren't retrofit lamps).

 

[Platonic Solid]

IES reports of same fixture housing and lens:
One with 2 F32T8 Fluorescent Lamps.
One with discrete LED and Driver (not a retrofit lamp so no reduction for extraneous fluorescent ballast):

Lithonia Lighting Fluorescent "WT" Series

Model: WT8 2 32
Total Lumens: 4552
Wattage: 46.7
L/W: 97

Lithonia Lighting LED "WL" Series

Model: WL4 40L LP840
Total Lumens: 4325
Wattage: 39.5
L/W: 109

 

[Platonic Solid]

Even better, the Cree Retrofit Lamp IES report actually defines the fixture used. So here we go:

One 2X4 troffer with 2 F32T8 Fluorescent Lamps.
One 2X4 troffer with 2 CREE LEDT8-48-22L-35K Lamps.

Lithonia Lighting Fluorescent "GT8" Series

Model: 2GT8 2 32 A12
Total Lumens: 4100
Wattage: 58
L/W: 71

Lithonia Lighting "GT8" Series enclosure w/CREE LEDT8

Model: LEDT8-48-22L-35K (Scroll down and click on IES 3500K)
Total Lumens: 3795
Wattage: 44
L/W: 86

 

[Denwood]

So if I interpret that correctly the ballast "cost" was about two watts for the set of Cree retrofit bulbs. If each is rated at 21 watts, and measured consumption is 44... Thanks for posting the IES data Platonic I wasn't aware of the IES site until now.

 

[Showkey]

Post 51 using watts per lumen then look at the cost per bulb in post 1 ........makes LED a tough sell. At least for average Joe's garage. ?????????

I know service life plays another part and color index yet another part.
I may not live long enough to see the service life savings

 

[Platonic Solid]

So if I interpret that correctly the ballast "cost" was about two watts for the set of Cree retrofit bulbs. If each is rated at 21 watts, and measured consumption is 44... Thanks for posting the IES data Platonic I wasn't aware of the IES site until now.

That IES site is worth spending some time playing with.
Here is the spec sheet for the CREE LEDT8P-48 which contains a different photometric report tested at CREE's photometric lab CESTL (Cree Engineering Services Testing Laboratory). I dug up the actual report and they are still using the GT8 Lithonia Troffer. The fluorescent ballast used in both reports is Sylvania QHE 2x32T8/UNV ISN-SC. Test Input Power: 120.23V, 0.360A, 43.23W. So the figures look a little better:

(2) LEDT8P-48-21L-35K (per CESTL REPORT# CESTL-PL04179-001) in Lithonia "GT8" 2X4 troffer
Total Lumens: 3918
Wattage: 43.2
L/W: 90.7

So CREE is publishing 2 different photometric reports. I wouldn't say they're significantly different, but it does raise a couple questions:

The previous report from their website defined the lamp model # as LEDT8-48-22L-35K. What's up with the "22L" designation as that doesn't seem to be an available lamp.

The spec sheet linked above has a "Watts vs. Lumens" chart which defines "LEDT8P-48-21L-35K HE". The photometric report is for "LEDT8P-48-21L-35K", so what does the "HE" stand for, High Efficiency? Is the "HE" lamp different from the non-HE?

All this is an attempt to answer you original question. Per the "Watt vs. Lumens" chart, it appears to be a 1.5W per lamp ballast loss, thus 3W per 2 lamp system or anywhere from 8% to 5% depending on ballast factor. Which is quite good, just not as good as 0% loss if no ballast was involved.

 

[Platonic Solid]

Post 51 using watts per lumen then look at the cost per bulb in post 1 ........makes LED a tough sell. At least for average Joe's garage. ?????????

I know service life plays another part and color index yet another part.
I may not live long enough to see the service life savings

I think the biggest selling point of LED lamps in our garages at this point is cold temperature starting. If you're looking for payback in your lifetime, you'll want to go for a higher L/W lamp such as the 140L/W Remphos. LED lamps are a lot like computer processors - it's never a good time to buy since the current model will be cheaper tomorrow and the newer model will be more efficient.

 

[Platonic Solid]

Before you whip out your light meter and measure your new lamps consider that in all my independent laboratory lumen maintenance tests thus far, LED output increases by 10% in the first 1000 hours of operation. I would go with the assumption that most IES reports are done using LED lamps that have been seasoned to achieve peak performance values.

 

[Denwood]

Actually with all of these tests you've posted..my motivation to test further just about disappeared. IES has already done the work! I won't use the Cree bulbs in hand in my shop (they'll go over to our Cinevate facility) however you've peaked my interest in a 90+ CRI low profile direct wire package, ideally under 1" in height. The Lithonia units with built in occupancy sensors are awesome. Our new Cinevate facility is 100% sensored lighting (ambience/heat/sound), and almost zero switches. I love the idea of holding individual lights off instead of banks as we do now. We do have entry/corridor T8 fixtures with a low power 2" bulb nestled between two 4ft ..and these have individual sensors. A competitively priced LED version (not sure on the Lithonia fixtures for price) where each light is programmed vs each bank would likely save even more power.

We did solar modeling for light gain before installing windows, so ambience sensing was important. Our one space with no windows (interior corridor) has three 16" sola-tubes which on a sunny day are bright enough to hold the corridor lighting off completely via the ambience programming on the sensors in that space. The best light is free

Our 9500 square foot building (renovated warehouse) uses a furred air-space assembly and a pile of thermal efficiencies (cheaper to heat than my house) however power use still hovers at 3000kwH/month. Last year over January (with an average temp of -19C) we spent a whopping $224 on natural gas to heat the 9500 sq/ft. Overall we're about 50% more efficient than a typical building. My goal is to get the business to net zero as far as the building goes...so roof mounted PV and efficiency go hand in hand to get there.

 

[cybrdyke]

All this is an attempt to answer you original question. Per the "Watt vs. Lumens" chart, it appears to be a 1.5W per lamp ballast loss, thus 3W per 2 lamp system or anywhere from 8% to 5% depending on ballast factor. Which is quite good, just not as good as 0% loss if no ballast was involved.

The lamp manufacturer's, the good ones anyway, publish two numbers..one is a "bare bulb" figure, the other is a "with ballast" figure, so you can deduce the consumption of the ballast.
It's not correct to say that there is 0% loss if no ballast was involved. There IS a ballast, called a driver, which is internal to the tube. Except in this case, the manufacturer does NOT tell you the consumption of each of the components separately. For all we know, the driver in the bypass tube uses even more watts than an existing ballast. Not that it matters anyway, because separately, the two components are worthless. But you always have to remember that there IS a ballast/driver to consider in the bypass tubes...and normally, it's the weakest link. Compare apples to apples.

I think the biggest selling point of LED lamps in our garages at this point is cold temperature starting. If you're looking for payback in your lifetime, you'll want to go for a higher L/W lamp such as the 140L/W Remphos. LED lamps are a lot like computer processors - it's never a good time to buy since the current model will be cheaper tomorrow and the newer model will be more efficient.

Agreed that in residential situations, which have short annual burn hours that it's hard to justify the cost of LED. Higher lumens per watt, all by itself, doesn't necessarily make a job payback better. In the case of the RemPhos tubes, 2400 lumens at 17 watts (141 lm/w) may not payback any better than a Cree tube 2100 lumens at 21w (100 lm/w). Plus, 2400 delivered lumens is overkill in a typical commercial retrofit situation. Most customers ask for the same amount or at least a similar amount of light that they currently have.
CD

 

[Platonic Solid]

... It's not correct to say that there is 0% loss if no ballast was involved. There IS a ballast, called a driver, which is internal to the tube. Except in this case, the manufacturer does NOT tell you the consumption of each of the components separately. For all we know, the driver in the bypass tube uses even more watts than an existing ballast. Not that it matters anyway, because separately, the two components are worthless. But you always have to remember that there IS a ballast/driver to consider in the bypass tubes...and normally, it's the weakest link. Compare apples to apples.

As you know, in the lighting business a Ballast = Fluorescent or HID Ballast and a Driver = LED Driver.
In that intended context, my statement is correct. There could be no loss from the Ballast if it didn't exist.

Agreed that in residential situations, which have short annual burn hours that it's hard to justify the cost of LED. Higher lumens per watt, all by itself, doesn't necessarily make a job payback better. In the case of the RemPhos tubes, 2400 lumens at 17 watts (141 lm/w) may not payback any better than a Cree tube 2100 lumens at 21w (100 lm/w). Plus, 2400 delivered lumens is overkill in a typical commercial retrofit situation. Most customers ask for the same amount or at least a similar amount of light that they currently have.

The primary purpose of these LED retrofit lamps is to save energy by retrofitting existing fluorescent luminaires. The CREE 2100 Lumen 21W lamp may be a desired choice for some fixtures with no diffusers (lenses), though diffusers are generally preferred for LED glare reduction. For existing fixtures with diffusers the RemPhos 2400 lumen 17W lamp is a better choice to offset aging diffuser lumen loss.

 

[Platonic Solid]

Actually with all of these tests you've posted..my motivation to test further just about disappeared. IES has already done the work!

Your personal experience still matters.

... you've peaked my interest in a 90+ CRI low profile direct wire package, ideally under 1" in height.

When do you want it? What features do you need? and how many? PM me.

Our new Cinevate facility is 100% sensored lighting (ambience/heat/sound), and almost zero switches. I love the idea of holding individual lights off instead of banks as we do now. We do have entry/corridor T8 fixtures with a low power 2" bulb nestled between two 4ft ..and these have individual sensors. A competitively priced LED version (not sure on the Lithonia fixtures for price) where each light is programmed vs each bank would likely save even more power.

We did solar modeling for light gain before installing windows, so ambience sensing was important. Our one space with no windows (interior corridor) has three 16" sola-tubes which on a sunny day are bright enough to hold the corridor lighting off completely via the ambience programming on the sensors in that space. The best light is free

Our 9500 square foot building (renovated warehouse) uses a furred air-space assembly and a pile of thermal efficiencies (cheaper to heat than my house) however power use still hovers at 3000kwH/month. Last year over January (with an average temp of -19C) we spent a whopping $224 on natural gas to heat the 9500 sq/ft. Overall we're about 50% more efficient than a typical building. My goal is to get the business to net zero as far as the building goes...so roof mounted PV and efficiency go hand in hand to get there.

Do you have any pictures of this space? It sounds incredible!

 

[Denwood]

Plat, pm sent. We took video during the entire build, from architect meetings until completion. Was thinking about a build thread here as I think the message is a good one. This is a bit of a video walk thru (minus the upper floor studio) during our Kickstarter/Morpheus Stabilzer campaign. If you're a Star Wars collector you'll like it even better. Ha.

https://vimeo.com/81443885

Efficiency on a budget.

 

[Platonic Solid]

Plat, pm sent. We took video during the entire build, from architect meetings until completion. Was thinking about a build thread here as I think the message is a good one. This is a bit of a video walk thru (minus the upper floor studio) during our Kickstarter/Morpheus Stabilzer campaign. If you're a Star Wars collector you'll like it even better. Ha.

https://vimeo.com/81443885

Efficiency on a budget.

interesting. Impressively smooth video too.

 

[Mike Michaels]

Thanks for the review.

I have 20 T8 bulbs installed in my unheated garage. Since I live in MN the lights remain on 24x7 from October to May. All the lights are on one circuit so there is no option of turning some off. Kids are coming and going all hours of the day/night and turning the lights off is a safety issue because there are no windows or other sources of light. The guy who wired the garage also did not put a switch by the service door.

Cold startup is the main reason for leaving them turned on. I replaced the bulbs in our coach lights this fall with LED and love the instant on compared to CFLs. For that reason I am considering replacing a few tubes in the garage with LED. Occasionally the lights do get turned off in winter and it would be helpful to have one or more banks with instant on capability.

And to defy convention - every single one of my ballasts failed in their first 12 months of operation after they were installed ten or so years ago. IIRC they were Sylvania (or Philips). Home Depot replaced them at no charge. Just this last year I had the first ballast failure since the original install/replace. Minor PitA.

 

[Denwood]

Mike, you might consider just installing a discrete motion sensor on your garage circuit, and add in a few LEDs. Then you'd have all lights off with no motion, and instant on with respect to the LED's. I added in a motion sensitive LED lamp in my garage which is powered from the roll up door operator circuit, so illuminates the garage when the lights are not on. This is to ensure that when the roll up door closes, the missus is not plunged in darkness It only works "at night" so is inactive with the lights on.

 

[Denwood]

Here's a few tests I ran based on the discussion raised by Platonic et al

The "test rig". Ha. You can see here that the fluorescent ballast uses about 3 watts with no lamp installed as reported by the EM100.

The following two photos were taken at exactly the same exposure...so the relative light brightness is as seen in the photos. The Cree LED uses exactly what you might expect as it's rated at 21 watts, and we know the ballast uses up 2-3 watts. 23 watts total.

Not sure how to explain this but the Philips TL950 with a 32 watt bulb is actually using 27 watts. Only five watts more than the Cree LED, with a 98 CRI output at 5000K.

 

[Platonic Solid]

GE-132-MV-N ballast has 0.88BF thus a regular F32T8 is being run at 28W (32 x 0.88). FYI that ballast does not meet new DOE standards and has been replaced with GE132MAX-G-N

 

[Denwood]

Thanks for that Platonic. I had to do some reading on "ballast factor" which explains Cree spec'ing the T8 led from 1700 to 2100 lumens depending on ballast.

 

[padstack]

I think the biggest selling point of LED lamps in our garages at this point is cold temperature starting. If you're looking for payback in your lifetime, you'll want to go for a higher L/W lamp such as the 140L/W Remphos. LED lamps are a lot like computer processors - it's never a good time to buy since the current model will be cheaper tomorrow and the newer model will be more efficient.

I seem to remember seeing a 150 lumens/watt option from energy focus at a recent show as well. Plug and play bulb just like the remphos ones mentioned here.

 

[Pantsfall_McFixit]

Seems like for a reliability standpoint it would be best to have just a bare fixture with 120v going to the pins, then a small driver on each lamp unit. Makes for cheaper fixtures. Of course the driver is usually what's made cheaply and kicks the bucket.

 

[fastjohnny]

Seems like for a reliability standpoint it would be best to have just a bare fixture with 120v going to the pins, then a small driver on each lamp unit. Makes for cheaper fixtures. Of course the driver is usually what's made cheaply and kicks the bucket.

That's what I'm doing with these:

http://www.ebay.com/itm/28152732742...49&var=580552670403&ssPageName=STRK:MEBIDX:IT

Very pleased so far.

 

[Denwood]

One of the gotcha's with low CRI light (like the ebay LEDs) is the fact that many of us do painting/staining etc. in the shop. For that reason, colour accuracy is a big deal IMHO..hence anything under 90 CRI and 5000K would rule out that light for shop use. If you have windows and like natural light to mix in with your overhead, again 5000K is what you want so the colour temperatures match.

If you're just looking for a warmer look (quite a few garages I see here are not working shops) to lounge, watch TV etc. then a warmer 3200K would make sense. In that case, CRI and light accuracy might be secondary. My 2c

 

[cybrdyke]

Even better, the Cree Retrofit Lamp IES report actually defines the fixture used. So here we go:

One 2X4 troffer with 2 F32T8 Fluorescent Lamps.
One 2X4 troffer with 2 CREE LEDT8-48-22L-35K Lamps.

Lithonia Lighting Fluorescent "GT8" Series

Model: 2GT8 2 32 A12
Total Lumens: 4658
Wattage: 58
L/W: 80

Lithonia Lighting "GT8" Series enclosure w/CREE LEDT8

Model: LEDT8-48-22L-35K (Scroll down and click on IES 3500K)
Total Lumens: 3795
Wattage: 44
L/W: 86

IES reports of same fixture housing and lens:
One with 2 F32T8 Fluorescent Lamps.
One with discrete LED and Driver (not a retrofit lamp so no reduction for extraneous fluorescent ballast):

Lithonia Lighting Fluorescent "WT" Series

Model: WT8 2 32
Total Lumens: 4552
Wattage: 46.7
L/W: 97

Lithonia Lighting LED "WL" Series

Model: WL4 40L LP840
Total Lumens: 4325
Wattage: 39.5
L/W: 109

These are not actually IES reports, but Acuity's photometric reports done in Visual lighting software. And they are filled with errors. Both of the reports in red above have gross errors in them and shouldn't be used for comparison. It's common for the manufacturers to have these mistakes in these types of reports. You have to be really diligent when going thru this type of report to find them.
CD

 

[cybrdyke]

Cool. I'm sure you'll find the numbers come out just as they should.
Happy New Year.
CD

Here's a few tests I ran....

The Cree LED uses exactly what you might expect as it's rated at 21 watts, and we know the ballast uses up 2-3 watts.

Denwood,
Confirmed! as they would say on Mythbusters. The electronic ballast will use anywhere from 1.5 to 3 watts per tube. As I stated previously, that's why they publish 2 numbers, a bare bulb number and a "with ballast" or "system" wattage.
Glad to see you got the right numbers in your test.

Oh...one other thing, completely off topic...since you're in the video game...have you heard anything about high frequency lighting playing havoc on HD video cameras? I am starting to see lighting products developed specifically for use in areas where HD TV cameras are being used.
CD

 

[cybrdyke]

Thanks for that Platonic. I had to do some reading on "ballast factor" which explains Cree spec'ing the T8 led from 1700 to 2100 lumens depending on ballast.

That's right. It's one advantage that the ballast ready tubes have over the bypass tubes. They will react to the BF of the existing ballast, giving you the proper amount of light. It will also save you a few extra watts, which is nice.
CD

 

[Denwood]

High frame rates (an iphone will do 240 fps!!) dictate high shutter speeds so lighting in some cases is a real issue with respect to strobing. Lights like our Profoto HMI have no such issues..but they are $6000 each...

Some resort to DC power and tungsten, or lights over 2000 watts (AC) which have fewer issues with the filament dimming at 60hz. Electronic ballast and flourescent are generally fine, but do have issues at higher frame rates and shutter speeds.

 

[Platonic Solid]

These are not actually IES reports, but Acuity's photometric reports done in Visual lighting software. And they are filled with errors. Both of the reports in red above have gross errors in them and shouldn't be used for comparison. It's common for the manufacturers to have these mistakes in these types of reports. You have to be really diligent when going thru this type of report to find them.
CD

Please provide specific details to back up claims.
I'm sure the small Acuity Brands Lighting Manufacturers (Lithonia, Holophane, Mark ...) would be very interested in knowing that their nationally accredited photometric labs are somehow substandard. You should contact them and inform them of their errors. The Visual lighting software is a generic IES viewer with some basic Coefficient of Utilization calculation abilities. That's like saying a digital photograph isn't a legitimate photograph because you opened it in Windows Paint instead of Photoshop.

 

[cybrdyke]

Remember when you posted this?

Being condescending does nothing for your credibility.

and then you post this??

I'm sure the small Acuity Brands Lighting Manufacturers (Lithonia, Holophane, Mark ...) would be very interested in knowing that their nationally accredited photometric labs are somehow substandard. .

Sorry, but these are NOT IES files. IES files are the property of the Illuminating Engineering Society and are used, mostly, to be dropped into a variety of lighting software or CAD programs, like this one from Acuity. What you are linking to is a datasheet created by Acuity that has PCD and ISO charts that were created by using the IES files. And the data listed on page 1 is typed in by a human, not created by the IES files. And humans make mistakes. Anybody that uses these files is aware that there are very often mistakes in this data and they look for it diligently before banking on it. And, yes, when a mistake is found, it's simply a matter of calling Acuity, or Cooper, or Hubbell, and telling them about it and they will correct it.

Hey, I'm just trying to point out some errors that you missed.
Get over yourself.

 

[Platonic Solid]

Cybrdyke, Just to be clear: You have not pointed out any errors, just stated that there are undefined errors.

 

[Platonic Solid]

Sorry, but these are NOT IES files. IES files are the property of the Illuminating Engineering Society and are used, mostly, to be dropped into a variety of lighting software or CAD programs, like this one from Acuity. What you are linking to is a datasheet created by Acuity that has PCD and ISO charts that were created by using the IES files. And the data listed on page 1 is typed in by a human, not created by the IES files. And humans make mistakes. Anybody that uses these files is aware that there are very often mistakes in this data and they look for it diligently before banking on it. And, yes, when a mistake is found, it's simply a matter of calling Acuity, or Cooper, or Hubbell, and telling them about it and they will correct it.

Sigh...
THESE ARE IES FILES. IES files are the property of the manufacturer who paid to have the test done. What I am linking to are IES files translated by a program designed to open it. All the data on page 1 is part of the IES file as put there by ACUITY BRANDS LIGHTING CONYERS LAB (nationally accredited photometric lab). Yes, I'm sure human interface was involved, as it is in all IES files.
Model: 2GT8 2 32 A12 is TEST# LTL7424
Model: WT8 2 32 is TEST# LTL21861

I'm through with this ******* contest. It's not worth my time or effort.

 

[e36jon]

Hey gang.

I just got my Cree LS4 fixtures and posted an un-boxing review thread here: http://www.garagejournal.com/forum/showthread.php?p=4558344&posted=1#post4558344

I'll try to keep to Denwoods quality standard when it comes to taking photos of them when I get them installed.

Cheers,

Jon

 

[Electric_Light]

lol, so did someone here comment over at designingwithLEDs as
led_master cmi? After search the comment there, I found this thread through Google.

http://www.designingwithleds.com/review-hands-cree-linear-led-t8-fluorescent-replacement-lamp/

The spectrum graph in that review uncovers the huge chunk missing between green and blue and all current production LED lights are missing anything shorter than the blue LED's spectrum which means that its completely missing UVA and violet spectrum. This prevents currently available LED lighting products from correctly rendering anything that have fluorescence or colors sensitive to proper spectral power distribution.

Go buy a box of the same kind of Triscuit I am showing you here. Everything about this item is spectrum sensitive. The crackers themselves, the picture of the crackers and the blue-green to golden yellow transition all show color shifts under different lighting. It's in the gamut range that's very difficult for my camera to capture or the monitor to display so I can't really display what I am seeing on the screen. I told the camera to white balance to a neutral gray card before taking each picture rather than using the "preset" daylight.

Visually very close to the way it looks under natural daylight. It should look bluish cyan. This is under 5000K color matching fluorescent lamps with CRI spec of 90. Under natural sunlight, incandescent, or wide spectrum fluorescent lighting you will see bright, deep, greenish marine blue. To me, almost sickeningly vibrant. Ironically, it looks just the opposite in these pictures. It looks almost the same under incandescent once you give your eyes the chance to adjust to color temperature difference.

Philips brand 5000K LED bulb. Note the peculiar color shift of printed crackers towards yellow. To my eyes, the green tint on the box was lost. It looks like shades of blue under a 5000K 90 CRI CREE/Eco Smart downlight as well. I believe it has to do with the interaction of the pigments with the spectral range between green and blue where LEDs have a spectral defect.

Pigments and colored transparent plastics showed the biggest skew. Things look very close between sunlight and colormatch 5000K fluorescent lamps, but not so with 5000K 90 CRI LED.
If everything shifted in color at the same time, it's just a matter of viewing pleasure. Unfortunately, they don't. The worst is when you try to get two sections to have the same color and they match up under defective light spectrum like LED, then have it show up as two different colors in the sun light.

It's convenient to play with words to pitch the absence of ultraviolet as a benefit for the LED industry simply because it's extremely difficult for affordable LED products to produce it. You'll find that different materials have different fluorescence. Light up with a black light. Soda bottles and such glow dim orange or yellowish. Some white plastics don't glow. Some glow brightly. So why does it matter? The binder or clear coat from different brands or batch may have different levels or color of fluorescence. This means that sources with inherent spectral defects like LEDs may show a good color match with rest of the vehicle, but you see a clear difference once you look at it in the sun light in the morning.

 

[Denwood]

Electric, your post comparing pictures of a general purpose LED low CRI vs high CRI fluorescent is pretty much stating the obvious, and would be the case with any low CRI lighting vs high CRI. If you have a comparison of the CREE T8 vs TL950 I'd be interested, otherwise, your post is has little to do with this thread.

The TL950 is a 98 CRI bulb. Comparing a 98 CRI source (in terms of color accuracy) to a 90 CRI light source and expecting them to be the same is unreasonable.

 

[Electric_Light]

Electric, your post comparing pictures of a general purpose LED low CRI vs high CRI fluorescent is pretty much stating the obvious, and would be the case with any low CRI lighting vs high CRI. If you have a comparison of the CREE T8 vs TL950 I'd be interested, otherwise, your post is has little to do with this thread.

The TL950 is a 98 CRI bulb. Comparing a 98 CRI source (in terms of color accuracy) to a 90 CRI light source and expecting them to be the same is unreasonable.

I have a CREE 5000K 90 CRI LED as well and the box claims "90 CRI excellent color quality". I described that the missing cyan/green under LED lighting was the same as with their general purpose 5000K LED bulb.

"Things look very close between sunlight and colormatch 5000K fluorescent lamps, but not so with 5000K 90 CRI LED". The F40T12/C50 lamps are also rated 90 CRI and actually extends into ultraviolet, but not as much as sun light.

"Full spectrum lamps in the visible range are designed to simulate the color range of sunlight. These include the sunshine, Chroma 50, and Chroma 75 fluorescent lamps. Visible light is found in the wavelength range of 400-700 nanometers and emits some UV, but the amounts are well below those found in daylight and sunlight, even when filtered through a window. These are not considered harmful. This is where normal, everyday use lamps (incandescent and fluorescent) fall. "