Color Rendition and Tint Comparison: Cree, Rebel, GDP, Nichia

selfbuilt

Flashaholic
Joined
May 27, 2006
Messages
6,936
Location
Canada
There is always a lot of discussion here on the relative color rendition quality of various LED emitter types (and the specific color temperature/tints within those models). I typically like sticking with objective test data that I can measure in a reliable way, but I thought I would try my hand at doing a relative subjective comparison between emitters and tints. :)

It is very hard to capture subjective visual impressions with a camera, but I've come up with the following method to try and facilitate the presentation:
  • For samples, I have picked lights with what I consider representative tints and emitters. Most of these are not specific known tint bins, but are based on my relative visual impressions compared to other lights in my collection (see Addendum for my tint bin estimates).
  • I've chosen as my test objects common colorful items a flashaholic may have on hand (including various common battery types, typical office supplies, and a few other familiar items ;).
  • Since the lights all differ in beam profiles, I've chosen to do a ceiling bounce comparison in a closet with white walls and ceiling, with my camera mounted on flexible portable tripod.
  • I've picked lights that can operate at relatively similar outputs, and equalized the exact output as close as possible on my camera (Canon Powershot S5). Specifically, I've set the white balance to daylight, locked the aperture at f2.7, and varied the shutter speed to capture equivalent overall exposures using my camera's built-in histogram feature (I manually set the shutter speed to generate as close to a perfectly centered histogram as possible for each light, thus equalizing the overall exposure).
  • For more details on the actual lights used, and my estimate of tint bin, see the Addendum at the end of this post
Click on the individual pics to bring up full-sized images - works best if your browser supports multiple tabs, so that you can easily switch between tabs and see the pics in a stationary way.















Again, click on the individual pics to bring up full-sized images.

Combination Comparisons:

To help you compare tints at a glance, below is an animated gif showing one "representative" tint for each of the four emitter types. Note the total colors had to be reduced to 256 for the GIF format, so you will see some hatching. I also had to reduce the overall dimensions, to keep the final file size down.

Animated GIF of the four emitter types:
Emitters.gif


I will leave it to you to draw your own conclusions about relative CRI (color rendition index) of the various emitters in this test. :whistle:

To be perfectly honest, personally I don't find that there's a huge difference among the various emitter samples. I think a lot of it really comes down to tint differences (which are great enough to obscure subjective color impressions). To help show you what I mean, here are the 6 Cree tints shown in an animated gif.

Animated Gif of the six Cree tint samples (from cool to warm):
Cree.gif


I don't know about you, but I find the reds and browns show up a bit better on the "warmer" tint Crees. The cooler tint Crees tend to look "whiter", and are better for distinguishing the blue and blacks. The differences within an emitter class thus seem greater to me than between these emitter brands. :shrug:

Bottom Line :

Note that I am not making any claims as to the scientific validity of my testing method above - I'm just trying to find a way to easily compare lights at a common overall exposure and removing any beam pattern effects. The specific settings of your graphics card and monitor will make a huge difference in how these images appear. Also note that the "daylight" white balance setting can be highly variable from one camera to the next, which would skew the results. So there are many possible confounds here - this is just meant to be an approximate comparison.

But at the end of the day, I don't really see a noticeable color rendition difference between any of these particular emitters. Subjectively, I am far more influenced by relative color tint and temperature differences, and I think there would have to be a pretty significant CRI difference for me to really notice it in use. :shrug: Of course, I hope that will change as newer LED emitters come out with significantly improved CRIs (e.g. I haven't tested the new Seoul "natural white" 4000K yet, or the new K2s). But since all the emitters here typically have a CRI >70 (with incans being ~100 CRI), I am not really expecting huge subjective improvements. But YMMV, and I welcome any observations you might have on the pics above.

Personally, I would just go with whatever tint and beam pattern you most prefer for a given task. For example, I personally like a premium to slightly warm "cool white" LED for general use. Outdoors, I like the slightly warm "cool white" tints for greens/browns - but don't like going as far as the "neutral white" Cree 5A tints. And my wife likes her cool tint Fenix L2T V2 Rebel - she uses it every morning to help her pick out matching black socks in her sock drawer (trust me, you can't do that with my warm-tinted lights). :laughing:

Hope you found the pics useful! :wave:

UPDATE: In case you are curious about the terms used here, please check out the CPF Welcome Mat for general background on tint and color temperature, among a great many other things.

------------------

UPDATE: although I didn't do outdoor shots for this comparison, I have previously compared the Cree neutral-white Fenix TK20 (likely 5A tint bin) to the standard Cree cool-white Fenix L2D (likely WD tint bin) here:

TK20.gif



------------------

Addendum:

In case you are curious, here are the actual lights I used, with my estimate of the tint bin:

Crees (from top to bottom above):
  • Zebralight H50 (cool white, cool bluish tint - I'd estimate WA/WB bin)
  • JetBeam Jet-I PRO IBS (cool white, cool purplish tint - I'd estimate WM bin)
  • Nitecore Defender Infinity (cool white, very premium white tint - I'd estimate WC bin)
  • Fenix L1D (cool white, warm yellowish tint - I'd estimate WD/WG bin)
  • Olight M20 (cool white, very warm yellow-greenish tint - confirmed WH bin)
  • Fenix TK20 (neutral white, likely 5A bin)
Rebels:
  • Fenix L1T V2 (cool white RB080, cool bluish tint - I'd estimate Y0 bin)
  • Fenix L1D (cool white RB100, warm yellowish tint - I'd estimate VN/UN bin)
GDP:
  • Nitecore D10 (cool white, overall yellowish tint but slight bluish corona)
Nichia:
  • AltusLumen PAD-L (cool white 0.5W, new emitters with premium white with slightly warm yellowish tint)

For help understanding the tint bin codes, CPF user DFiorentino has compiled graphs for many of the major emitters here.
 
Last edited:

NairB

Newly Enlightened
Joined
Jan 30, 2006
Messages
18
What a fantastic post.

I am a newbie and reading through the forum regarding all the different tints, this post has let me seen easily what you have all been discussing regarding the tints.

I can't decide which I prefer looking at the cool whites (cool) or the warm (WH)....??

I think the cool wins for me :sssh:
 

Haz

Enlightened
Joined
Apr 14, 2005
Messages
919
Location
Sydney, Australia
Thanks for another great post. I like this very much and will be useful in helping me decide my next purchase
 

selfbuilt

Flashaholic
Joined
May 27, 2006
Messages
6,936
Location
Canada
Glad you are enjoying the comparisons. :D

I've just added some additional background detail with my estimate of tint bins for each of the samples. Also included is an update with an exterior shot showing the Cree cool-white (i.e. WD) vs neutral-white (i.e. 5A) from an early thread.

Cheers!
 

NoFair

Flashlight Enthusiast
Joined
Dec 22, 2004
Messages
1,556
Location
Norway
Very nice review!

I strongly prefer the warm tints and love the Cree 5A ones. Works a lot better outdoors and just as well indoors IMHO.

Sverre
 

Owen

Flashlight Enthusiast
Joined
Feb 14, 2002
Messages
2,047
Location
AL
Awesome comparison:thumbsup:
That warm Rebel, and .5w Nichia look pretty nice.
All of my current user LED lights(except 5mms) have 5A Cree emitters, except for one, and that one is in a box getting shipped out to have a mod that includes a Q3 5A.
I couldn't believe how bad all my LEDs looked in actual use compared to a Malkoff M60W(arm) when I first got one, and can hardly stand to use other types now.
 

OceanView

Enlightened
Joined
Nov 12, 2006
Messages
416
I couldn't believe how bad all my LEDs looked in actual use compared to a Malkoff M60W(arm) when I first got one, and can hardly stand to use other types now.
I know what you mean. LED light always bugged me, and not until I got my first Cree 5A tint did I really feel comfortable with how LED light looks. But primarily, I think it's just a preference thing.

I know that the neutral and warm tints are getting lot's of buzz now, but maybe it's mostly the folks like me, who have been dissatisfied with cool whites, who have mostly been propelling the buzz. I suspect that there are still plenty of CPFers out there who prefer how cool whites look to their eyes. Plus, they're still the most efficient, and for lumen junkies, cool whites are still the way to go, and visually, cool white does appear brighter.

Once the warm/neutral tint boom settles down, I wonder what the mix of sales of Gene's cool versus warm Malkoff drop-ins becomes going forward? I wouldn't be surprised if cool whites win out over the long term.
 

clg0159

Enlightened
Joined
Aug 3, 2007
Messages
446
Location
North Texas
Nice post! Thanks for taking the time to do the comparison:thumbsup: I really like the tint on that AltusLumen PAD-L. A truly interesting new light!
 

McGizmo

Flashaholic
Joined
May 1, 2002
Messages
17,287
Location
Maui
Selfbuilt,
Great work! :thumbsup: I suspect in most cases and provided the CRIa of the LED is perhaps 70 or above, there is enough information available for color recognition and only in some problem colors (purples for instance) will the differences between light sources be obvious. I am not sure that high CRI itself will ever become a big criteria or significant goal in most light design parameters. It seems that color temperature and tint alone will be the deciding factors in most cases. Should you find yourself doing more analysis of color, I highly recommend you consider one of the color checker cards. They make a very informative target and one that others can compare to your images if they have a card as well.
 

selfbuilt

Flashaholic
Joined
May 27, 2006
Messages
6,936
Location
Canada
Nice post! Thanks for taking the time to do the comparison:thumbsup: I really like the tint on that AltusLumen PAD-L. A truly interesting new light!
Agreed - although the PAD-L looks slightly warm to my eyes, the camera seems to capture it as one of the most neutrally-white (?) of all the "cool white" series. Of course, that could just be the bias of my camera's daylight white balance, or my own monitor/graphics card settings ... but I personally like the PAD-L's tint and floody beam profile.

Selfbuilt,
Great work! :thumbsup: I suspect in most cases and provided the CRIa of the LED is perhaps 70 or above, there is enough information available for color recognition and only in some problem colors (purples for instance) will the differences between light sources be obvious. I am not sure that high CRI itself will ever become a big criteria or significant goal in most light design parameters. It seems that color temperature and tint alone will be the deciding factors in most cases. Should you find yourself doing more analysis of color, I highly recommend you consider one of the color checker cards. They make a very informative target and one that others can compare to your images if they have a card as well.
Thanks for the comments McGizmo. I really appreciate you offering your perspective, given your extensive experience in building lights with various emitters. And I think you are bang on with your assessment. :)

Good tip on the color checker card. The problem with this sort of random object analysis (aside from lacking a proper color distribution, of course) is that I'll never be able to replicate it without redoing all the lights again. :laughing: I'll see if I can pick a card up at a local photographic supply store ... it will definitely come in handy for any future comparisons.

BTW, my intent here was somewhat "proof-of-concept" - to see if the testing and presentation method has value. The ceiling bounce and camera exposure equalization seems have worked fairly well ... but I'd appreciate any and all input from users on how to improve the testing regimen.
 

Cheapskate

Enlightened
Joined
Feb 24, 2008
Messages
346
Location
Ireland
Good tip on the color checker card. The problem with this sort of random object analysis (aside from lacking a proper color distribution, of course) is that I'll never be able to replicate it without redoing all the lights again. :laughing: I'll see if I can pick a card up at a local photographic supply store ... it will definitely come in handy for any future comparisons.

Paint brochures or catalogues are free and can provide a wide gamut of colours.
 

divine

Flashlight Enthusiast
CPF Supporter
Joined
Sep 30, 2007
Messages
1,829
Location
Virginia
Great work, Selfbuilt.

There are some holes to fill in on that list. K2 being one. In my experience, the K2's give a pretty good rendering.
 
Top