Which tint or colour temperature?

[robert.t]

I am curious to know which colour temperature people prefer and, most importantly, why?

You may give more than one answer, but then you also need to give more than one reason.

I haven't completely made my mind up yet, but I'm edging towards neutral white because it gives the most natural colour rendition.

[Joeymt3]

I prefer the neutral for the same reason you mentioned.

[AnAppleSnail]

Tint and color temperature are different.

Here is a picture.



That thin dotted line is "pure untinted" CCT, low in the bottom-right, high in the top-left. The colors are exaggerated tints. The curved squares are Cree color bins. My favorite is a very high CRI at high (6000-7000K) CCT for color-based tasks, and a very high CRI at a medium CCT (4000-6000K) for relaxed tasks. I prefer minimal tint, but if I must have one, I like slight yellow or blue tints, unless I am cooking. Then red, but never green.

The short version is, "The only full description of an LEDs output is its spectral power distribution (A list of each Wavelength and the ratio of output at that wavelength). We model light with measurements to approximate this shape. One example is "CCT, CRI, and Tint." That is, you may have a Cree LED at 5000K CCT in the 5A (Slightly peach yellow) color bin. That is easier to describe than a list of CIE color-space coordinate ranges (The corners of the boxes in the Cree Tint Map), and much easier than a long list of SPD components.

[robert.t]

Tint and color temperature are different.

Good point and that's a very useful reference image, thanks. Of course one problem I find is not always having the full specification for the LED in question, so I might say that I like the output of such and such a light, and I know that it is "neutral white" but I don't necessarily know the exact tint or colour temperature.

[mcnair55]

I prefer what ever comes with the light and have no interest at all as long as the light serves its purpose of illumination.If i buy a light from a trade store they never ask me just hand me the box.

[ven]

I am sure a 3c is neutral robert,or one of them that is classed as a NW....

I am still learning myself,so have a from a warm 4c tint(very warm) to cool white 1a.
Not over struck too much on the NW being honest,the creamy white is not my preferred although the jury is still out.
Warm i like,but gives an artificial colour as everything being exaggerated.....
Cool i find gives better detail and i find better for throw
Nichia 219 is a nice tint to me
Looking forward to the tn35vn mt g2 which i believe is around 5500k

Convoy to name one brand offers various tint options,(also good lights too for the money),so an affordable way to experiment with different tints,some you will prefer,ones you do then can be chosen in more expensive lights or for specific aplications,maybe the greens of forests/woods/fields to mountains etc etc.

But as its said it can be a tint lottery,some will have green/blue etc colours if white wall hunting,i prefer outside as everything changes,what on a wall may look poor,outside is great.............but i am sure i am not the only one who checks their tints on white walls

I mentioned in another thread about rach(the boss) showing me the difference and prefers cool and why as gives better detail. Also better colour rendition and not artificial,showed me going through around 5 or so of my lights with different tints.

[Fireclaw18]

5000k neutral white is a pretty nice tint. For Cree emitters its the 3C tint.

I also like 4000k and 4500k tint. Anything without green in it. I hate green tint.

[Etsu]

I haven't completely made my mind up yet, but I'm edging towards neutral white because it gives the most natural colour rendition.

The most natural colour rendition would probably be had from a cool-white high-CRI light. Sunlight is about 5500K, which is in the cool-white range. And if you're in the shade, or it's a cloudy day, the colour temperature could be 8000K or higher. That is the colour temperature in which our eyes evolved to see colour. It's not warm white or neutral white, it's cool white.

However, if your goal is not a natural colour, but rather a light that emphasizes reds and yellows, then you're better off with a warmer LED like a neutral or warm white. I think that is what most people prefer.

There's also another reason to choose a warmer LED. Warmer LEDs are usually higher CRI, just by the manufacturing process I suppose. A higher CRI is really your goal in colour, so most people choose a warmer LED because that's the easiest way to get a high CRI.

If you're doing colour work in a job, you're probably going to use daylight florescent lights, which are kind of like cool-white florescents with a high CRI. Personally, I find them too blue for regular lighting, but they are great for colour rendition.

Personally, I like the tint of the Nichia 219 neutral white. High CRI, and it really makes reds and browns pop.

[robert.t]

The most natural colour rendition would probably be had from a cool-white high-CRI light. Sunlight is about 5500K, which is in the cool-white range. And if you're in the shade, or it's a cloudy day, the colour temperature could be 8000K or higher. That is the colour temperature in which our eyes evolved to see colour. It's not warm white or neutral white, it's cool white.

I'm not so sure this is true, but part of the reason for this thread is to find out which people prefer and why. The important difference between LED light and sunlight is that sunlight is full-spectrum white light. That is, it contains lots of wavelengths at similar strengths. LED light is a much narrower spectrum designed to fool our eyes into thinking it's the same "colour" of light.

To me, CW tends to have a definite blue or green tint. WW looks more "natural" in the sense that it looks more like incandescent or firelight and gives the sort of light we're used to seeing in the dark - it isn't so harsh, but colours are clearly more "orangey" than they should be. There's also less total light output. NW is somewhere between these extremes.

One of the reasons incandescent lights look "better" for a given brightness is that they produce a fuller spectrum of light than an LED does. However as you note, the colour temperature is still very different to sunlight.

[hiuintahs]

I like LEDs without any bluish or greenish tint. I also don't like the warm (yellowish) tints. I'd say that my favorite is around 4750 to 5500k and maybe that is on the neutral side of cool white. Now that I have purchased LEDs with a bin spec, I'd say that I like the border area between 2A, 2d, 3A, 3D and 2B, 2C, 3B, 3C. I've actually bought 2 leds with the same bin spec and found one to be slightly different than the other as each of those bin squares is not a finite value but within a range. So most of the time I accept what I get unless its a big disappointment.

[mcnair55]

I bought a new light today over the phone,the guy asked do i want cool white or neutral and i replied stick whichever comes to hand in the box first makes no difference to me as long as it serves its intended job.

[ven]

I bought a new light today over the phone,the guy asked do i want cool white or neutral and i replied stick whichever comes to hand in the box first makes no difference to me as long as it serves its intended job.

But what is its intended job,one may serve the other better
What light has mr mcnair purchased?

[cland72]

I prefer the Nichia 219 over any other LED/tint I've seen. That puts me in the 4500k fan club, I suppose.

[kbuzbee]

Overall, it's a Nichia 219, so, neutral w/ high CRI. I do like a warmer tint as well. Feels "homey" but I seem to grab a Nichia based light more often than not. Today it's a D25C but there are others .

I know cool white throws better / is more efficient all other things being equal but this D25C is plenty bright enough for 99.5% of the things I need.

Ken

[RetroTechie]

Nichia 219 has a fan club all of its own.

[Poppy]

I bought a new light today over the phone,the guy asked do i want cool white or neutral and i replied stick whichever comes to hand in the box first makes no difference to me as long as it serves its intended job.

I know my friend that you are just trying to bring home the point that you don't really care about tint selection, but if you look at the CREE XML Data sheet, the neutral white bins deliver on average about 12% fewer lumens than the U2 Cool white bin. Unless you are getting a really cheap/inexpensive light, the cool white light will be a U2 bin. So if the purpose of the light is just to see more in the dark, you might get more bang for the buck with the cool white.

On the other hand, if your purpose is to have a power failure light to ceiling bounce and just sit around and chat with the misses, then you may prefer the softer white of the neutral white.

The difference in performance is not earth shattering, but it might be worth considering next time you order a light that you have the option of choosing a tint/color temperature.

Good point and that's a very useful reference image, thanks. Of course one problem I find is not always having the full specification for the LED in question, so I might say that I like the output of such and such a light, and I know that it is "neutral white" but I don't necessarily know the exact tint or colour temperature.

robert.t,
I fully agree. The only lights that I have that I know the tint specification is my convoy S2 where you can choose amoung a number of different tints when you order it. And a few of the P60 dropins that I built.
Unfortunately it seems that the reflector and how hard an LED is driven also affects what the beam looks like.
I'll have to play later tonight, or tomorrow morning with darkness and dark adapted eyes to see which ones I prefer.

[mcnair55]

But what is its intended job,one may serve the other better
What light has mr mcnair purchased?

One of your recommendations.

[mcnair55]

I know my friend that you are just trying to bring home the point that you don't really care about tint selection, but if you look at the CREE XML Data sheet, the neutral white bins deliver on average about 12% fewer lumens than the U2 Cool white bin. Unless you are getting a really cheap/inexpensive light, the cool white light will be a U2 bin. So if the purpose of the light is just to see more in the dark, you might get more bang for the buck with the cool white.

On the other hand, if your purpose is to have a power failure light to ceiling bounce and just sit around and chat with the misses, then you may prefer the softer white of the neutral white.

The difference in performance is not earth shattering, but it might be worth considering next time you order a light that you have the option of choosing a tint/color temperature.


robert.t,
I fully agree. The only lights that I have that I know the tint specification is my convoy S2 where you can choose amoung a number of different tints when you order it. And a few of the P60 dropins that I built.
Unfortunately it seems that the reflector and how hard an LED is driven also affects what the beam looks like.
I'll have to play later tonight, or tomorrow morning with darkness and dark adapted eyes to see which ones I prefer.

Mr Poppy,

I must be lucky as the guy said he was putting cool white in my box.

[ven]

One of your recommendations.

[Wiggle]

Depends on the application but personally I find anything in the 4000K (warm-ish neutral) to 6500K (cool white) can be best. For general EDC I find a nice cooler white is what I will usually gravitate towards. I find warmer temperatures need a higher drive level for me to appreciate them and EDC is typically for low lumen outputs.

For use in the woods or anything with lots of rich colour a temperature closer to 4000K can be very nice and provide good contrast between shades of brown and green. For throwers, I just got a K50vn dedomed (approx 5000K) that seems almost purely neutral and it seems to scatter a little less than the cool white lights I have.

Nature seems to respond well to warmer and neutral lights while urban environments are better handled with a slightly cooler light. Just IMO of course

[Swede74]

Tint and color temperature are different.

Here is a picture.



That thin dotted line is "pure untinted" CCT, low in the bottom-right, high in the top-left. The colors are exaggerated tints. The curved squares are Cree color bins. My favorite is a very high CRI at high (6000-7000K) CCT for color-based tasks, and a very high CRI at a medium CCT (4000-6000K) for relaxed tasks. I prefer minimal tint, but if I must have one, I like slight yellow or blue tints, unless I am cooking. Then red, but never green.

The short version is, "The only full description of an LEDs output is its spectral power distribution (A list of each Wavelength and the ratio of output at that wavelength). We model light with measurements to approximate this shape. One example is "CCT, CRI, and Tint." That is, you may have a Cree LED at 5000K CCT in the 5A (Slightly peach yellow) color bin. That is easier to describe than a list of CIE color-space coordinate ranges (The corners of the boxes in the Cree Tint Map), and much easier than a long list of SPD components.

I am still having trouble understanding the difference between colour temperature and tint. I think I know what a tint snob is, and I believe I became one the same day I received my Zebralight SC62d but it would be nice to be able to be able to better understand, and perhaps even contribute to threads like this one. When I first joined CPF I was confused by the terms "lumen" and "lux", but after a while, that mystery was solved. I'm hoping the colour temperature / tint confusion will clear up too, and was wondering: is it possible to use sound as an analogy?

If you strum a violin string and a guitar string, both tuned to 440 Hz (A) they don't sound very alike, even though they vibrate at the same frequency. Now, if the two instruments were two flashlights instead, would it make sense to say they have the same colour temperature but different tints?

A second example: if you press the same string against different frets along the neck of a guitar, could you say the tones generated have different colour temperatures but the same tint?

[TEEJ]

The reason it makes no sense is that, ironically, "color temperature" ONLY applies to WHITE LIGHT.

So all of the babble about temperature is essentially meaningless. ..except for some consensus as to concept to misuse.

The color can be a frequency. ..as the wavelength in nanometers defines the color.

If looking at a sunrise or sunset. ..you might notice that dawn and dusk lighting have different characteristics. ..With the morning perhaps being a cooler light, and the sunset being a warmer light.

As the sun's light is the same...The angle it's coming in at can change our perception of it.

This concept is what the °K is supposed to be describing.

[robert.t]

As the sun's light is the same...The angle it's coming in at can change our perception of it.

This concept is what the °K is supposed to be describing.

My understanding is that it's supposed to indicate the bias within a full light spectrum towards either the blue end or the red end. The sun always puts out the same spectrum of white light, and this changes as it rises and sets not because our perception changes, but because of the way the light refracts within the atmosphere. The fact that different wavelengths refract at different angles is why rainbows exist and why the sky is generally blue. As the angle between the sun and the ground changes, the spectrum of light reaching the ground changes, so more red gets through. That's also why the sky starts to turn red at sunset.

This only makes really makes sense if you have a graph of the light energy at different wavelengths, which for white light will give you a continuous line (the area under which indicates the total light output). For artificial light, you'll actually get a number of blips at certain very specific wavelengths. This means you have mix of colours, not true white light. As you say, colour temperature is not strictly speaking applicable in this case, but I've never entirely understood the concept anyway, it's just a huge oversimplification of how light actually works - it just seems to have become popular recently on TVs, monitors and suchlike, which are incapable of emitting true white light anyway.

For an incandescent light, the spectral graph would be more or less white, but with a significant bump near the red end of the spectrum, hence the low colour temperature compared to daylight. LEDs however, put out a mixture of colours in much tighter frequency bands. I'm not exactly sure what a frequency graph looks like for a typical white LED, but I've always assumed they work more-or-less like the RGB mixture that produces white in TVs and monitors, since this is optimised to fool our eyes into seeing white. However, the absorption and reflection characteristics of materials don't work the same way, so the colours reflected back by such artificial light can look rather off.

This might also be why some people report that incandescent lights are better for depth perception. I'm not entirely sure why that is, other than that we perceive depth through both parallax (mainly stereo vision) and "shape from shading". The latter may be affected by artificial light spectra. Apparently, women rely more on SFS and men rely mostly on parallax.

I had a quick search about for spectrum graphs, but couldn't find any reliable sources, so if anyone has some for some common LEDs, please post them. It would really help to illustrate the true difference between them. Hue/tint; colour temperature: these are all simplifications anyway.

[TEEJ]

My understanding is that it's supposed to indicate the bias within a full light spectrum towards either the blue end or the red end. The sun always puts out the same spectrum of white light, and this changes as it rises and sets not because our perception changes, but because of the way the light refracts within the atmosphere. The fact that different wavelengths refract at different angles is why rainbows exist and why the sky is generally blue. As the angle between the sun and the ground changes, the spectrum of light reaching the ground changes, so more red gets through. That's also why the sky starts to turn red at sunset.

This only makes really makes sense if you have a graph of the light energy at different wavelengths, which for white light will give you a continuous line (the area under which indicates the total light output). For artificial light, you'll actually get a number of blips at certain very specific wavelengths. This means you have mix of colours, not true white light. As you say, colour temperature is not strictly speaking applicable in this case, but I've never entirely understood the concept anyway, it's just a huge oversimplification of how light actually works - it just seems to have become popular recently on TVs, monitors and suchlike, which are incapable of emitting true white light anyway.

For an incandescent light, the spectral graph would be more or less white, but with a significant bump near the red end of the spectrum, hence the low colour temperature compared to daylight. LEDs however, put out a mixture of colours in much tighter frequency bands. I'm not exactly sure what a frequency graph looks like for a typical white LED, but I've always assumed they work more-or-less like the RGB mixture that produces white in TVs and monitors, since this is optimised to fool our eyes into seeing white. However, the absorption and reflection characteristics of materials don't work the same way, so the colours reflected back by such artificial light can look rather off.

This might also be why some people report that incandescent lights are better for depth perception. I'm not entirely sure why that is, other than that we perceive depth through both parallax (mainly stereo vision) and "shape from shading". The latter may be affected by artificial light spectra. Apparently, women rely more on SFS and men rely mostly on parallax.

I had a quick search about for spectrum graphs, but couldn't find any reliable sources, so if anyone has some for some common LEDs, please post them. It would really help to illustrate the true difference between them. Hue/tint; colour temperature: these are all simplifications anyway.

Correct, I think you're getting it.

There is no such thing as monochromatic white light.

Just as a prism breaks the suns white light into its component parts, the angle the rays are coming in at dusk or dawn changes the way the light looks to us. (The distance through the atmosphere is proportional to the shmootz it passes through on the way in, and, am and pm levels of dust and moisture, etc, will differ, and so forth)

Now, if you think of the sun as a big 'ol flashlight, the color temperature of the sun's rays is "white light", as its a blend of the spectrum, and what that giant flashlight is emitting doesn't change just because part of its blocked by a planet...only the perceived colors are shifted...on the planet.

So, you'll make yourself crazy trying to find a real answer, as there really isn't one.

Its essentially boiled down to the wavelength is the color....and, the temperature is another unrelated way of attempting to say the light is more like sunset or sunrise lighting.

Tint is essentially COLOR, as it the beam is more yellow, or more green or more blue, etc.

This would be like getting a pail of white base paint at the paint store, and, then handing them a color chip, and they TINT it to match the COLOR you want.



You might THEN say....hmmm, I want a COOLER tint, or, a WARMER tint.

And they add some yellow or red, etc, to adjust it accordingly. (Same BASE color, but, a cooler or warmer version)


If you said to raise it by 500º K, they would not know how to do that...even though the CONCEPT is the same....there's just no conversion factor for them to use.

[Swede74]

The reason it makes no sense is that, ironically, "color temperature" ONLY applies to WHITE LIGHT.
So all of the babble about temperature is essentially meaningless. ..except for some consensus as to concept to misuse.
The color can be a frequency. ..as the wavelength in nanometers defines the color.

As you say, colour temperature is not strictly speaking applicable in this case, but I've never entirely understood the concept anyway, it's just a huge oversimplification of how light actually works - it just seems to have become popular recently on TVs, monitors and suchlike, which are incapable of emitting true white light anyway.

It seems there is not only more to it than meets the eye...

If looking at a sunrise or sunset. ..you might notice that dawn and dusk lighting have different characteristics. ..With the morning perhaps being a cooler light, and the sunset being a warmer light.
As the sun's light is the same...The angle it's coming in at can change our perception of it.

As the angle between the sun and the ground changes, the spectrum of light reaching the ground changes, so more red gets through. That's also why the sky starts to turn red at sunset.

...it is also a matter of how it meets the eye.

I must admit I'm still a bit confused (and wishing I had stayed in school longer) but I thank you for responding. Hopefully I will get a better grasp of things eventually.

[TEEJ]

It seems there is not only more to it than meets the eye...





...it is also a matter of how it meets the eye.

I must admit I'm still a bit confused (and wishing I had stayed in school longer) but I thank you for responding. Hopefully I will get a better grasp of things eventually.

Re-read my last post...I posted it while you were typing your's...it might add some support to your quest.

[robert.t]

If you strum a violin string and a guitar string, both tuned to 440 Hz (A) they don't sound very alike, even though they vibrate at the same frequency. Now, if the two instruments were two flashlights instead, would it make sense to say they have the same colour temperature but different tints?

This is in fact quite similar to what is going on. A string tuned to 440Hz actually produces lots of harmonics - 440Hz will be the lowest frequency, which is what you get if you imagine holding both ends of the string and swinging it so the middle moves up and down 440 times per second. The second harmonic is what you get if the string is vibrating at double that rate, in which case each half of the string is moving up and down rapidly and the middle isn't moving. The 3rd harmonic is where the string is divided into 3, vibrating at 3 times the original frequency. And so on.

The reason the two strings sound different is the different relative power of each of the harmonics (along with different attack and decay characteristics of the volume). The important point here is that the sound isn't just one frequency, it's a spectrum of lots of frequencies. That is what gives you the timbre of the sound.

Light is exactly the same. One colour (or hue) is one specific frequency of light. Many colours are made up of a mixture of frequencies, which is why there are many colours we can see, in print for instance, that we can't produce on an RGB monitor.

However, the terminology for light and sound is completely different. There is no direct analogue for timbre for light. A hue is a pure colour (one frequency), so is most analogous to a pure tone of 440Hz, which of course sounds nothing like either a guitar nor a violin. Tint and shade attempt to describe something more like the timbre of a colour.

Tint is a slightly imprecise term, that just means a colour that is "off white" by mixing white with some hue. The HSV colour system is based on tints, and tries to describe all colours as a combination of hue (the base colour), saturation (how much white is added) and value (the brightness). However this is also a simplification that is unable to describe all visible colours.

Shades are not applicable to light because that's what you get when you mix a hue with black rather than white, which for a light would just give you the same colour.

Colour temperature essentially means the exact same thing as tint, except that as has been pointed out it's really a technical term used in relation to sunlight (black body radiation, to be even more pedantic about it), that has been (mis)appropriated for use elsewhere. Really the only difference is that a tint would maybe be a slight bump (or dip) in a spectral graph, whereas colour temperature would be a shift of the entire graph, higher on one side and lower on the other - a bit like a wave versus a tide.

[Etsu]

The way the white-balance on a camera describes tint is roughly as follows:

- The color temperature is the relative amounts of red vs blue light,

- The tint is the relative amounts of green vs magenta light.

Thus, it might register a given light as having a color temp of 5000K, with a tint offset of +2 green. It can describe most "white" just using those two variables.

It's not an exact way of describing tint, but it's the simplest I know. For most lighting situations, it works pretty well, certainly if the light is natural or an incandescent source. It has a bit more trouble with florescent lighting, because the spectrum has large peaks and valleys.

[kbuzbee]

I had a quick search about for spectrum graphs, but couldn't find any reliable sources, so if anyone has some for some common LEDs, please post them.

I don't feel comfortable posting it but Jason has a nice graph of the Nichia 219 on the Darksucks site. Here's a link to it:

http://darksucks.com/store_BetaQR.html

Ken

[TEEJ]

This is in fact quite similar to what is going on. A string tuned to 440Hz actually produces lots of harmonics - 440Hz will be the lowest frequency, which is what you get if you imagine holding both ends of the string and swinging it so the middle moves up and down 440 times per second. The second harmonic is what you get if the string is vibrating at double that rate, in which case each half of the string is moving up and down rapidly and the middle isn't moving. The 3rd harmonic is where the string is divided into 3, vibrating at 3 times the original frequency. And so on.

The reason the two strings sound different is the different relative power of each of the harmonics (along with different attack and decay characteristics of the volume). The important point here is that the sound isn't just one frequency, it's a spectrum of lots of frequencies. That is what gives you the timbre of the sound.

Light is exactly the same. One colour (or hue) is one specific frequency of light. Many colours are made up of a mixture of frequencies, which is why there are many colours we can see, in print for instance, that we can't produce on an RGB monitor.

However, the terminology for light and sound is completely different. There is no direct analogue for timbre for light. A hue is a pure colour (one frequency), so is most analogous to a pure tone of 440Hz, which of course sounds nothing like either a guitar nor a violin. Tint and shade attempt to describe something more like the timbre of a colour.

Tint is a slightly imprecise term, that just means a colour that is "off white" by mixing white with some hue. The HSV colour system is based on tints, and tries to describe all colours as a combination of hue (the base colour), saturation (how much white is added) and value (the brightness). However this is also a simplification that is unable to describe all visible colours.

Shades are not applicable to light because that's what you get when you mix a hue with black rather than white, which for a light would just give you the same colour.

Colour temperature essentially means the exact same thing as tint, except that as has been pointed out it's really a technical term used in relation to sunlight (black body radiation, to be even more pedantic about it), that has been (mis)appropriated for use elsewhere. Really the only difference is that a tint would maybe be a slight bump (or dip) in a spectral graph, whereas colour temperature would be a shift of the entire graph, higher on one side and lower on the other - a bit like a wave versus a tide.

I think this is a good representation.