Color Rendering Test - LED vs. Incandescent

[RichS]

This is a re-do of a previous color rendering post where I wanted to see if the LED or the incandescent light would do a better job of rendering Easter Eggs my daughter had just colored. As some of you pointed out, the test was flawed because I hadn't set the manual white balance separately for each light.

****UPDATE****
​

After gaining more insights from posts received on this thread, I've learned some important info about doing comparison shots. As it turns out, I did take the correct approach with my first comparison thread after all by using a common white balance setting (daylight) that won't allow the camera to adjust or "correct" the colors that are rendered from each light source.

So by re-doing my color rendering test using manual white balance for each light and having my camera "adjust" to each light source, I essentially forced my camera to "make it look right", by enhancing the blues, greens, reds, etc. as it saw the light source had deficiencies in specific areas of the color spectrum.

So that being said, I must re-post the original results. HERE are the ORIGINAL color comparison shots I took with each light, with the white balance set to "daylight" for both shots. In these shots it is very easy to see the characteristics of each type of light and how it affects the color, contrast, depth, dimension, and warmth of each image. When I look at the images, I can see that it will definitely be each person's personal preference as to which light source they prefer, as they are each so unique. As for me, I prefer the warmth and depth of the incandescent image, even though a lot of the reds, pinks and oranges are harder to separate. The LED image just looks a little cold and flat to me, but it make it easier to distinguish the colors in my opinion.


WE Raider (Incandescent) / WE Sniper (Q5 LED)

****White Balance Set to Daylight for Both Lights****

Wolf-Eyest Sniper Q5 (LED)

Wolf-Eyes Raider (incandescent)

Here are my "learning experience" shots (thank you CPF'rs) where I set the white balance manually for each light. It is very easy to see in the manual white balance shots how the camera corrected each image to the point that they look very close to each other, even though they are very different light sources.

****White Balance Set Separately for Each Light Using White Paper****

Wolf-Eyest Sniper Q5 (LED)

Wolf-Eyes Raider (incandescent)

Conclusions:

This time the results were very different. Although the pictures show both lights did very well, the incandescent light rendered the colors much truer to my eye. The LED colors seemed a little off, almost fluorescent for some of the colors like the pinks, while other colors look a little dull. Also as was noticed before, the incandescent light seemed to show the depth and dimension of the eggs more, as opposed to the LED light which caused the image to look slightly more flat.

So, in the previous test I had the LED light in the winner's circle, but as you correctly pointed out the test was flawed. In this test which I feel is a much more fair representation of each light's ability to render colors, the incandescent light wins in my book - hands down.

-Rich

 

[SilverFox]

Hello Rich,

Would you happen to have a flash on your camera for a third perspective...?

Tom

 

[Norm]

the incandescent light seemed to show the depth and dimension of the eggs more, as opposed to the LED light which caused the image to look slightly more flat.
-Rich

To me the second picture has more depth not because of the light source but but the different angle of the light, notice the deeper shadows on the incandescent picture. It would be interesting to see both pictures shot with the light source in identical positions. I'm sure it's the lack of shadows making the LED image look flat.
Norm

 

[GreySave]

Nice job. Glad the experiment paid off. I tend to rely on the automatic white balance myself more than I should. Might have to try a few manual shots.

 

[MikeSalt]

Any chance of a shot illuminated purely by sunlight? I agree, the incandescent is marginally better, but the Q5 is certainly doing a good job.

I have three lights with the Cree Q5 emitter, with three different tints. One is slightly blue, one is spot-on white and the other is slightly yellow. Of the three, the slightly yellow tint illuminates colour depth just as nicely as a well-driven incandescent. I'm just waiting for a Lumens Factory HO-E1R bulb to verify this test against something a little stronger.

 

[RichS]

Hello Rich,

Would you happen to have a flash on your camera for a third perspective...?

Tom

I do Tom, but unfortunately I don't have the eggs anymore. I pitched them after taking the pics last night since they were over a week old.

To me the second picture has more depth not because of the light source but but the different angle of the light, notice the deeper shadows on the incandescent picture. It would be interesting to see both pictures shot with the light source in identical positions. I'm sure it's the lack of shadows making the LED image look flat.
Norm

I see what you are saying about the slightly larger shadows in the incandescent picture, but I don't think this is making that big of a difference. To me, in the incandescent picture the eggs have more contrast between the edge of the egg and the shadow, making the eggs "pop" more and appear more 3D. The shadows in the LED image aren't much smaller, but they are a little more difficult to see because of the lack of contrast between the egg edge and the shadow.

I did my best to hold the lights in the exact same place (upper left), but it I was holding it myself so they may be just slightly off. I think it was close enough to get a pretty good comparison though.

 

[RichS]

Nice job. Glad the experiment paid off. I tend to rely on the automatic white balance myself more than I should. Might have to try a few manual shots.

Thanks GreySave. I appreciate the expert advise which helped me to get to a successful comparison here. This time the outcome was what I originally suspected with the incandescent coming out on top. No doubt the LED can't be beat for unreal runtime and brightness for size, but there's no denying the pure illumination of a good incandescent light outdoors, and this simple color rendering test sheds some light on why.

 

[RichS]

Any chance of a shot illuminated purely by sunlight? I agree, the incandescent is marginally better, but the Q5 is certainly doing a good job.

I have three lights with the Cree Q5 emitter, with three different tints. One is slightly blue, one is spot-on white and the other is slightly yellow. Of the three, the slightly yellow tint illuminates colour depth just as nicely as a well-driven incandescent. I'm just waiting for a Lumens Factory HO-E1R bulb to verify this test against something a little stronger.

I agree Mike, the warmer LED tints do come close, but I've not seen one do quite as good as any incandescent outdoors. The best LED tint I have is a the P3D Rebel 100, and it is excellent. But it still doesn't quite match my nicer incandescents though. Of course, in the end it all comes down to each individual's personal preference and everyone sees things a little differently.

 

[SilverFox]

Hello Rich,

I totally understand... Oh well, there is always next year for a repeat test.

Tom

 

[Cheapskate]

I personally think incandescents have poorer colour rendition. I did a test for another thread using a maglite vs a couple LED torches. The camera is the one in my phone, A SE K750i. The white balance was set to daylight:

Adjusting the white balance manually between shots is surely a fudge, as it will make the output of an incandescent seem whiter than it actually is.

 

[270winchester]

are you seriously comparing 70 dollar LED lights to a 10 dollar light with 60 year old technology?

I personally think incandescents have poorer colour rendition. I did a test for another thread using a maglite vs a couple LED torches. The camera is the one in my phone, A SE K750i. The white balance was set to daylight:

Adjusting the white balance manually between shots is surely a fudge, as it will make the output of an incandescent seem whiter than it actually is.

 

[RichS]

I personally think incandescents have poorer colour rendition. I did a test for another thread using a maglite vs a couple LED torches. The camera is the one in my phone, A SE K750i. The white balance was set to daylight:


Adjusting the white balance manually between shots is surely a fudge, as it will make the output of an incandescent seem whiter than it actually is.

Really?? Have you taken your lights outdoors? Ok, here are a couple of other shots I took with no manual light balance setting (before I knew how to do it). Still think the LEDs do better? Ignore the output differences of each light as this doesn't effect the color rendering.

DBS 3SD Q5 SMO

Mag w/ WA 1185 bulb (incandescent)

CL1H 3SD Q5 OP

SL TL-3 Carly 1499

Regarding setting the white balance manually (fudging it, as you call it), the real photographers on this forum are the ones who instructed me to manually set the white balance if I wanted to get the true representation of each light's color rendering. This gives the camera the information it needs so it knows what white looks like under each type of specific lighting. It can then more accurately reproduce the rest of the colors appropriately based upon what white looks like under that light. It only makes sense, as setting it to one specific white balance essentially forces the camera to guess at what the color balance should be, or to add a color "filter" to the shots based on the preset you use such as "daylight".

BTW, if you want to see this same comparison done with the same white balance used, take a look at this one where I used the "daylight" white balance setting for both shots. Does the LED look better in these shots?? http://www.candlepowerforums.com/vb/showthread.php?t=192952

In my opinion, the negative effects were very obvious when I took my first shots, using the popular "daylight" balance to take them. Because when I got them loaded onto my PC, I didn't remember either of the shots looking that way to my eye.

 

[BVH]

Speaking of color rendition, anyone tried video bias lighting behind their TV? Read this thread.

http://www.candlepowerforums.com/vb/showthread.php?p=2420808#post2420808

 

[L.E.D.]

The new extremely warm Inova lights will come very close to incans, if not exceed them in some cases due to the lottery.

When those new nano-tube non-phosphor emitters come out with nearly 100 in CRI, they will make incans look like a joke, maybe even HID's, regarding "color rendering". Hopefully.

 

[RichS]

The new extremely warm Inova lights will come very close to incans, if not exceed them in some cases due to the lottery.

When those new nano-tube non-phosphor emitters come out with nearly 100 in CRI, they will make incans look like a joke, maybe even HID's, regarding "color rendering". Hopefully.

I agree it will be a fine day when those new 100 CRI emitters are released - I'll be the first in line! Until then though, I'll have a couple of good incandescents within arm's reach.

 

[gadget_lover]

IIRC, setting the white balance using a white object is an exersize that allows the camera to alter it's settings to match the light source. If there is not enough blue, for instance, the camera boosts the blue till the whitest thing in the frame is "white".

Our brains do this automatically, which is why you can see a white newspaper as white using a 60 watt incandescent that is really pretty yellow. You only really notice the difference when you have two lighting conditions in close proximity where you can see both at once. For example, two LED lights against a white wall.

In short, re-balancing between pictures mimics what your brain does.

White LEDs are generally deficient in green wavelengths, but make up for that by having the proper mix of blue and yellow light. That's why the incandescent almost always looks better against bushes and leaves. The incan has more green in it's spectrum.

[ Note: SOmething seemed wrong about that, so I looked at the spectrum charts on LEDMUSEUM's web site (http://ledmuseum.home.att.net/ledleft.htm) and see that the typical white LED has lots of blue and some green with a bit of yellow but with fairly little red. ]

That's my understanding...


Daniel
(BTW, Blue LIGHT and yellow LIGHT make white LIGHT. Blue PAINT and yellow PAINT makes green PAINT.)

 

[Outdoors Fanatic]

Damn! Those incans are kicking some serious LED's butts.

Great topic! Thanks for sharing.

Cheers

 

[MichaelW]

The new extremely warm Inova lights will come very close to incans, if not exceed them in some cases due to the lottery.

I really like my '08 T2-mp.
It is the warmest LED light I have. It looks yellow in comparison to the others.

But then compared to a 4D Mag, it still looks blue. and once again 4D with 3d bulb.

I estimate the color temperature to be 4000K. (to the nearest 500K)
Hopefully within six months, Inova will move to the warmest of Neutral-whites.

 

[RichS]

IIRC, setting the white balance using a white object is an exersize that allows the camera to alter it's settings to match the light source. If there is not enough blue, for instance, the camera boosts the blue till the whitest thing in the frame is "white".

Our brains do this automatically, which is why you can see a white newspaper as white using a 60 watt incandescent that is really pretty yellow. You only really notice the difference when you have two lighting conditions in close proximity where you can see both at once. For example, two LED lights against a white wall.

In short, re-balancing between pictures mimics what your brain does.

White LEDs are generally deficient in green wavelengths, but make up for that by having the proper mix of blue and yellow light. That's why the incandescent almost always looks better against bushes and leaves. The incan has more green in it's spectrum.


That's my understanding...


Daniel
(BTW, Blue LIGHT and yellow LIGHT make white LIGHT. Blue PAINT and yellow PAINT makes green PAINT.)

Daniel,

Thank you - the "light" went on after reading your post. Not immediately, it took me a while of mulling it over before it did, but it finally did...

So by re-doing my color rendering test using manual white balance for each light and having my camera "adjust" to each light source, I essentially forced my camera to "make it look right", by enhancing the blues, greens, reds, etc. as it saw the light source had deficiencies in specific areas of the color spectrum. Yeah, this is the right thing to do in normal photography, but not when the purpose is to capture each light's raw characteristics.

Now it makes sense to me why it is common practice here on the forum to set the white balance to one specific setting to do comparisons between different lights. And it only makes sense to use the "daylight" setting as sunlight is the gold standard to compare all other illumination sources against. Since sunlight renders colors so perfectly, the camera doesn't try to make adjustments by boosting colors to correct for the light source.

I must now also "eat crow" and apologize to Cheapskate for completely dismissing his point about "fudging" by taking this approach. You were right, ok...

So that being said, I must re-post the original results. HERE are the ORIGINAL color comparison shots I took with each light, with the white balance set to "daylight" for both shots. In these shots it is very easy to see the characteristics of each type of light and how it affects the color, contrast, depth, dimension, and warmth of each image. When I look at the images, I can see that it will definitely be each person's personal preference as to which light source they prefer, as they are each so unique. As for me, I prefer the warmth and depth of the incandescent image, even though a lot of the reds, pinks and oranges are harder to separate. The LED image just looks a little cold and flat to me, but it make it easier to distinguish the colors in my opinion.

Also with the reminder Daniel gave about how our brains adjust to this "white balance" automatically (I read this somewhere else before), I now understand why the results seen in these pictures didn't look this extreme to my eye as I was taking the picture.

I learn something new around here every day!

Wolf-Eyes Sniper LED (Q5)

WE Raider (Incandescent)

-Rich

 

[RichS]

I really like my '08 T2-mp.
It is the warmest LED light I have. It looks yellow in comparison to the others.

But then compared to a 4D Mag, it still looks blue. and once again 4D with 3d bulb.

I estimate the color temperature to be 4000K. (to the nearest 500K)
Hopefully within six months, Inova will move to the warmest of Neutral-whites.

Just curious, how warm are they compared to a Rebel 100 - do you have one to compare it to? I have a P3D Rebel 100 and it is very warm, but I seriously doubt it is close to 4000K. I need to get me one of those new Inovas...I use to have the older X01, X03, and X5 but sold 'em all after discovering CPF and Fenixes...then Derees...Olights..Surefires..Wolf-Eyes..........