Sharon Hicks Fine Art Home About the Artist Contact the Artist Artwork Portfolio


WARM OR COOL? Ultramarine Blue vs Thalo Blue ...

Many years ago (we won't say exactly HOW many, but it's been a good long while) I set out on what has become a lifelong journey of exploring colour theory. As a young adult I exhausted every book on the subject at the local libraries wherever I lived, and I never stopped. I guess you could say it's been a never-ending passion, this thirst for colour knowledge. Nowadays of course, my library of choice is Google, and there are endless numbers of articles out there on various theories about how colour 'works'. And there are discrepancies out there too. Lots of them.

One thing I've been coming across lately, and I can't really say for sure just when I first noticed this, are articles from various sources which state that Ultramarine Blue is a warm blue, while Thalo Blue is a cool blue.

This flies in the face of what I studied about colour temperature 'way back when'. I learned years ago that Ultramarine is cool, while Thalo is warm. When did this shift in viewpoint 'out there' occur? And why?

Those are the questions I've been pondering for the past few weeks, while I was busy updating notes for a workshop I was teaching on Basic Colour Theory. I'm the type of person who just “has to know”, and so I set about to try and figure out just where this shift may have stemmed from. And I think I now may have the answer.


To determine Colour Temperature, most artists refer to the well-known Colour Wheel, which arranges colours in a circle. We're all familiar with the 3 Primaries (R – Y – B) spaced at equal distances around the wheel, with Secondaries (O – G – V) positioned between them, and so on. Various versions of this basic colour wheel are commonly used by the majority of artists nowadays, and I believe this just might be where the confusion may have arisen over the relative temperatures of Ultramarine Blue and Thalo Blue.

The colours on the Red-orange-yellow side of the colour wheel are classed as Warm colours, while those on the violet – blue – green side are classed as Cool colours. That division is widely accepted, and is based in part on psychological connections our brains make when viewing certain colours. Red, orange and yellow are all seen as 'fire' colours, and our brains interpret them as warm. Violet and blue suggest icy shadows on snow, while green suggests the cool grass beneath our feet on a hot summer day. Seems straightforward enough, right? There's a lot more to it than that, of course, but this division of warm and cool is accepted as accurate.

When we compare two different blues, in this case Ultramarine Blue and Thalo Blue, with the generic blue of the colour wheel, we can see that Ultramarine leans slightly toward the violet side of Blue, while Thalo leans toward the green side of Blue. So that places Ultramarine closer to Violet, while Thalo is closer to Green.

On this `standard` colour wheel, Violet is positioned next to Red, and Red is in the Warm zone. Therefore, if we use the colour wheel as our only reference, we might conclude that a Blue which is closer to Violet is also closer to Red, and therefore is warmer than a Blue which is closer to Green, which is further away from Red. So we might surmise from this that Thalo Blue is cooler than Ultramarine. This, I believe, is what a lot of people are doing.



However, let`s take a closer look at this colour wheel. Where does it come from?

To fully understand the way the Colour Wheel 'works', we first need to take a look at the Colour Spectrum. What is this?


White light is comprised of different individual colour wavelengths, which our eyes and brains interpret as different colours. When this white light is broken down into its separate colour bands, as it is in a rainbow, we can see the different colours arranged in order of their wavelengths.

The colours in this simplified version of the Spectrum are arranged according to their relative wavelengths, with Red having the longest and Violet the shortest. Colour Temperature applies to this spectrum as well, with the colours on the Red end being warm, while those on the Violet end being cool. This is based on science, with the colours arranged according to measureable differences.

Now, let's bend the ends of the spectrum upwards (hypothetically of course), until the two ends join together. This forms a circle, which is what we call the Colour Wheel.

The point where Red meets Violet on the wheel actually indicates where the two ends of the spectrum meet.

Violet is not 'warm', even though it sits beside Red on the wheel - it is still the coolest colour in the spectrum, while Red is still the warmest colour. Those properties don't change simply because we've turned the Spectrum bar into a circle, or a wheel.

This appears to be where the notion arises that Ultramarine Blue is warm, while Thalo Blue is cool. If we use just the colour wheel, without considering its origins and underlying meanings, that would be the logical conclusion.


* * *

Now let's look just at the cool end of the spectrum - from Green through Blue to Violet.

We can see that Blue is situated between Green on the left, and Violet on the right. From the basic spectrum we learned that Violet is the coolest colour and Red is the warmest. Therefore, since Green is closer to the Red end than Blue is, then Green would be considered warmer than Blue. Conversely, Violet would be considered cooler than Blue.


Next, let's compare Ultramarine Blue and Thalo Blue with this portion of the spectrum.

First we will see that Thalo Blue has slight leanings toward Green, while Ultramarine Blue has slight leanings toward Violet.


Therefore Thalo would be located on the 'green end of blue', while Ultramarine would be located on the 'violet end of blue'.

Continuing on this line of logic, which is based on scientifically measurable properties, we can see that Ultramarine is closer to Violet, while Thalo Blue is closer to Green.


Violet is cooler than Green.


Therefore it follows that Ultramarine Blue is cooler than Thalo Blue.


* * *

In summary, then - for accurate colour temperature comparisons, it's essential to go back to the source - the Colour Spectrum. That will give us the truest reference.


* * * * *

(Sharon Hicks is an artist who resides in Sackville NB. )


14 Responses to WARM OR COOL? Ultramarine Blue vs Thalo Blue ...


What an interesting article!

One thing that struck me right away, was the appearance of your two color samples at the beginning of the article.

As near as I can tell from "computer screen samples" they are very close in value, and yet which one advances (characteristic of warm) and which one recedes (cool characteristic)?

Wow, what we know to be technically true, and yet what the eye perceives - color is a tricky game to play!


Sharon Hicks
Hi Karen ... good observation - it's difficult to portray the colours 'exactly' on the screen ... I actually wrestled with that problem for quite awhile - scanning and re-scanning my samples until I had it as close as I could make it appear to the real thing. But even at that, the colours in the samples shown are not really 'true', and it will of course vary even further on readers' monitors - there are so many different degrees of colour resolution out there when it comes to computer screens.

The best source of course is for each person to lay out their own colours, in the flesh so to speak, and examine them with a critical eye. That's what I'm always encouraging artists to do when I teach classes on anything to do with colour. I've been making colour charts and studying 'real' colours for as far back as I can remember, and I haven't stopped yet - there's always something more to check into ...

btw - in the real samples of Ultramarine and Thalo, it becomes quite obvious that Thalo advances and Ultramarine recedes ...

cheers, and thanks for commenting :)

David Briggs
The convention regarding violet as warmer than blue is actually the older of the two, but both go back a long way (early 19th century vs early 20th century). The opposing argument is that warm/cool is a psychological association of hue perceptions, not physical wavelengths, and violet hue perceptions contain a component of redness. Historically there have been a lot of different conventions on which hues are warm, cool or neither, and I don't think it's really possible to say that any one of them is "truest". However I'd be very interested to know where you learnt the "violet=coolest" theory, and anywhere else that you know it is being or has been taught.


My two-cents worth:

Your conclusion that Thalo Blue is warmer is likely correct, but I would you a different way to reach that conclusion.

Although your straight-line spectrum theory would appear to work for blues, zooming out into the big-picture, using that logic would also imply that more 'fiery' colours like Vermillion, Orange-Reds and Yellow are cooler than 'true' reds like Crimson Alizarin.


My understanding of colour is two-fold; colours in the visible-light spectrum are different in nature to pigment-based colour (as in when you paint) in the following ways:

- The primary colours of visible light are comprised of Red, Green and Blue, while the counterpart for pigment is Magenta, Yellow and Cyan.

- The primary colours of visible light total-up into white light, while the same for pigment would give a dirty brown, requiring that true blacks and grey be a separate pigment.

- Magenta does not naturally occur in visible light (it has to be either reflected off magenta pigment or artificially generated on-screen), and 'violet' in the light spectrum is not the same as 'violet' or purple in pigment because the former is actually reaching the edge of visibility to the human eye (ultraviolet), while the latter naturally starts to re-integrate itself into red.


One other factor to consider in 'warm' vs 'cool' values is 'intrinsic' brightness (although it has less to do with light and more with how the human eye perceives light). You have mentioned this in an earlier comment about colours that 'advance' and 'recede'. Green, for example, has a higher 'intrinsic' brightness than Blue, thus when comparing Green and Blue colours of equal value, Green tends to 'glow' or 'advance' more naturally - this would be the basis by which I conclude that Thalo blue is warmer.

Not to mention that that Green is so close to Yellow, which has the highest 'intrinsic' brightness of all, and by all accounts is functionally the 'warmest', even though it is obviously far-away from red! Certainly, a mere photo of the golden sun-scorched sand of the Sahara can be said to appear warmer than the reddest apple, haha...


I would also mention that dealing with Blue values is a unique problem. No one has trouble deciding that Crimson or Burgundy is cooler than Vermillion Red; no one argues that Lime Green is warmer than Teal, but every argues when it comes to blue, and in reality this is because the human eye simply cannot understand blue quite as well. The human eye has evolved to focus more on red and green for survival reasons - red indicates ripe fruit or blood, while green indicates healthy foliage or toxic animals. From a survival standpoint, blue is frankly a bit useless and at best esoteric (hence the spiritual connotations it is often linked to).

Fun digression: The naming of colours in the evolution of a cultures language also reflects this set of priorities: After 'light' and 'dark', Red is one of the first colours which are named, whereas Greens and Blues are among the last, with Purple waaaaay at the back. In fact, some cultures today still do not have separate terms for blue and green today. The Chinese language, for example, defined its 'Green' from its 'Blue' with separate words relatively late in its civilization, and the original umbrella term is still used today, albeit ambiguously and often loosely translated into to 'Azure' or even simply 'Fresh Colour'.


Your plight with explaining Ultramarine as cooler online is that by scanning your pigment, you are essentially trying to convert your pigment-based colour into light-based colour on a screen, and *everything* on a screen is produced by shining light out of it, so you have a field-day going out of your way to try and compensate for this, nevermind the additional issue of variability in colour display across screens!

I am a jeweller by profession and I witness a similar struggle with retailers selling blue sapphire and blue gemstone jewellery online in particular, compared to other things like ruby. Customers are almost always disappointed with the physical turnout because the computer interprets sapphire's deep blue colouring as a superlative of its light-based blue, and ramps it up for all its worth!

If you have Adobe-Photoshop or similar (I use a free software called Gimp which is almost exactly the same), you can manually reduce the brightness to better match.

Sharon Hicks
vintage scissors - I really appreciate your taking the time to provide the in-depth colour notes ... these various findings are all familiar to me, as I have spent quite a bit of time investigating various different approaches to colour. In the classes I teach, however, I find that the simpler I can explain a concept, the more likely the class participants will be to put it into use. I find that by using the spectrum analogy it gets the point across quite well, without having to go into the finer points of light-vs-pigment, comparative vibrancies, etc. It's kind of the K.I.S.S. principle for teaching colour to those who want a simple quick lesson. I must say I really enjoyed your ramble though - right up my alley :)


Hahaha, yes, I did go on a roll there (they happen more frequently and go on for longer now that I am older)!

Yes I agree that K.I.S.S. is best for a pragmatic, results-driven approach, especially for learners. If they are interested enough then it is up to them to delve further into theory on their own...

Victor LaPorte
I have been struggling with this very question related to developing a color pallet for oil painting.

This is important to me because Im trying to mix my own secondary colors from 6 primary colors. 3 warm and 3 cool of each primary.

The more important detail I'm trying to clarify is you can be unknowingly adding a third color to a mix when your intent is to only be mixing two colors. The result is a duller mixed color. For example ultramarine has a bit of red in it. So if I use it with yellow like lemon yellow it will make a duller green than if I had mixed it with thalo blue.

All of this makes my head hurt. The more I think I know the less I know:)

Sharon Hicks
Hi Victor ... I feel your pain LOL ... sorting out warm and cool colours, and colour mixing in general, can give one a gigantic headache.

Fact is, very few pigments would be what you could call 'pure' colours ... by having a warm and cool version of each primary, we can work around that tendency.

To get a 'good' green, you would select a yellow which is already on the greeish side, and a blue which is already on the greenish side. If you mix Lemon Yellow and Thalo Blue, for example, it will give you a 'truer' green than if you use an orangey yellow, or a purplish blue.

Mind you, those other combinations can also produce some really interesting greens. With just 2 blues (one warm, one cool) and 2 yellows (one warm, one cool), you can mix a wide range of interesting green tones. Mix each of the yellows with each of the blues, in different strengths - ie: more yellow and less blue, more blue and less yellow ... etc ...

Then you can do the same exercise with warm and cool red + warm and cool blue, and then of course warm and cool red + warm and cool yellow ... the possibilities are endless.

THEN you can start experimenting with mixing some of the secondary colours together, to get a wide array of greys and browns ... it's a fascinating topic which you can play with forever.

Have fun with it :)

I had a lot of problem with this. Ultramarine looked cooler than cobalt blue or phthalo blue to my eyes and I was reading the opposite in a lot of places. Finally I stopped thinking in terms of warmth or coolness and just looked to see if the blue was closer to green or violet. That's all I needed for help in mixing vibrant colors and in choosing which colors to harmonize or contrast. See Linda Kemp, Simplifying Color and Design, pages 40-445.

Sharon Hicks
Cheerful Monk - sounds like surrendering to your instinct was the best way to go. I think we have an 'innate' sense of colour. Choosing which blue is closer to green or closer to violet is really an instinctive use of the colour spectrum, I would say. Basing colour decisions on just the colour wheel can be deceiving, UNLESS one understands where the colour wheel comes from. Once you can visualize the colours in a straight line spectrum, it is far easier to determine relative coolness and warmth of any colour pair. Thanks for posting.

Robert Jones
I don't really agree with this, and if it shows anything it may be that the temperature we apply to colours is misleading in itself. Pthalocyanine blue is certainly a very strong colour, and one that doesn't obviously recede - it would therefore not meet all the criteria for 'cool' colours. On the other hand, straight Pthalo Blue contains no red, whereas Ultramarine does: it's not that Ultramarine is near Violet; it actually carries a proportion of red in its makeup. That's what makes it a relatively warm colour, but I don't think much of this makes sense unless you judge a colour in relation to others when you're actually painting with them, which we more or less do by instinct (as you say). The problem with your approach is that when it comes to mixing, it runs into problems - eg, no colour is going to be warmed by the addition of Pthalo Blue; quite the reverse. Violet, similarly, is only cool in relation to, eg, Vermilion. Lay it alongside Terre Verte and then tell me which is the warmer.

I respect the way you've worked this out, but it doesn't work for me.

David Briggs
Very good point, Robert. Not only does the colour violet contain a component of red perceptually - violet and violet-blue pigments do not just reflect the wavelengths seen as those colours, but have a peak in the red part of the spectrum as well, which may be why this argument based on the spectrum is very uncommon (though not unprecedented). More common arguments for the minority position (i.e. that violet/blue-violet is coolest) are based on the colour wheel and the assumptions that yellow or orange-yellow is the warmest colour and that red is neither warm nor cool.

joan forman
Excellent explanation... love your very clear visuals. I'm a retired teacher not art. I think you are a great teacher. :-)

Edit My
Artist Websites by FineArtStudioOnline
Mobile Site | iPhone Site | Regular Site