Blue, Blue, My Love Is Azure Blue - and happily some of our paints are too 😀

[John]

I decided to do a little experiment this afternoon to see if I could prove a pet theory, which is that Humbrol inadvertently matched the wrong standard when they brought out their Authentic Azure Blue in the 1960s, and the wrong shade persists in the current Humbrol 157.

When BS381 was first issued in 1930, Colour No 4 was listed as Azure. It's a strong dark blue, lighter than Oxford Blue and not a great deal different from the current Roundel Blue. Subsequently, the Royal Aircraft Establishment created the bright blue Desert Scheme colour we are familiar with as Azure Blue - similar name but completely different shade! It had occurred to me to wonder if Humbrol had matched the wrong Azure, not realising there were 2 shades with more or less the same name. There is another theory that Humbrol matched their paint against Light Mediterranean Blue, rather than Azure Blue.

I looked out the oldest tin of Azure Blue I could find in my paint box, which is an early version of 157 rather that the authentic HB13. and since it is a "blue stripe" tin probably dates to when 157 was introduced to the Standard range, which Richard Humm dates at 1988 in his list on the Airfix Collectors Forum. This colour was brushed out and compared to the chips in the RAF Museum book:



It looks nothing like Azure Blue, as we know, but it's also not that close to LMB either. LMB is bluer and less grey than 157. 157 is also markedly darker and greyer than the current BS381c-104 Azure, so it looks like both these theories are wrong at least based on this particular incarnation of 157\HB13. We're still no wiser when it comes to what Humbrol were thinking when they introduced this shade!

Comparing 157 to FS595, the nearest match is 35164:



Turning to the chip of AM Azure Blue, it is actually a very hard colour to match. There's nothing in RAL or BS381c that comes close, while FS595 gives you 35231 but it's darker and greyer than AB:



Even a Pantone deck only gave approximate matches, and they were from the Violet end of the scale. To my eye at least, AB has always had a violet caste to it so that's not too surprising:



Humbrol isn't the only company who have struggled with Azure Blue. Xtracolor and Xtracrylic are actually different shades, and neither is a precise match:





I'll track down some other attempts at AB and see how they compare, both with each other and with the sample chip.

John

[JosephLalor]

The only other possibilities that occur to me are Deep Sky or Dark Mediterranean Blue, but I think they would be darker than 157.

Joseph

[stevehnz]

John, I'm not sure that you are correct in assuming that early 157 equates to HB13. This is a pic I put up in another thread re Azure Blues, the underside of the old Rareplanes Vindicator on the left is painted with HB13, & is pretty close, to my eye , to the RAF museum colour chip. The wee spit on the right has the mix from the Humbrol colour system as found in the IPMS charts. Why oh why can't Humbrol be persuaded to get it right, they must, after all, have had a heap of evidence thrust upon them by now as to how hopeless 157 is? I don't know what the tin the HB13 came out of looked like but it would have been probably about 30+ years ago.

Steve.

[Prop Duster]

I'm am always fascinated and amazed at the knowledge and effort modelers will go to, for the goal of matching the correct colors.

I have read IPMS missives and many threads, on many forums about colors.

Overall my feeling is that the subject is an open ended one and will not be resolved for a long time. While I am not a stickler for color, I would and do like to use the color that best represents the model I am working on.

So Thank You John for adding to our "tribal knowledge" .

But, as noted, there have been many bits of evidence compiled and presumably, passed on to Humbrol. So, is there any evidence the Humbrol has ever, explained their choices? or what original data they used, to decide?

re:

" ... in my paint box, which is an early version of 157 rather that the authentic HB13. and since it is a "blue stripe" tin probably dates to when 157 was introduced to the Standard range,..."

It is obvious that the color swatch's shown here are at odds, but I would respectively ask, could the original, admittedly old Humbrol,

perhaps have changed in some way its color, hue or tint over these approximately 20 plus years?, thus making it some different from the newer examples.

Regardless John, thanks for an enlightening addition to our data. and happy modeling.

[Mike]

Velly interesting Mr John. Have you tried Lifecolor's version yet? UA098 is their Azure Blue - I can send you a brushed out sample if you like?

[Test Graham]

Interesting work, John, but I can add to the comments that at least some Humbrol tins have changed hues over the years, and there were several cases of changes between the original Authentics and those produced when they were merged into the main line. I don't recall Azure Blue as being one of those mentioned, but perhaps this is something your work may be hinting at?

[Nick Millman]

I am surprised by the appearance of the Xtracolor sample as I had always thought - without actually checking - that Xtracolor's Azure Blue was one of the closest to the original. I used it on a Kittyhawk and found that UV exposure (of the model) caused a colour shift to a much "cooler", greener blue. This has resulted in the model now having a patchwork appearance underneath! On the face of it the photo of the sample doesn't look like it though. Mine is a very old tin - so has the Xtracolor Azure Blue also changed over time?

The FS 595B comparisons are interesting too, because although 35231 is cited as being closest (also by Geoff Thomas) in pure colour science terms 35240 is actually closer. Neither can be considered close matches though. In Munsell Azure Blue is another PB - Purple Blue - with a red pigment to it, actually 4.1 PB 6.2/6.0 so your perception of a "violet caste" is spot on. Geoff Thomas wrote that it was 5 PB 6.5/6.0. The difference between the two is 2.66 where <2.0 = a close match, with Geoff's being paler, perhaps the result of his measuring an actual faded sample.

To complete the math 35240 is 3.6 PB 5.6/5.5 and 35231 is 4.8 PB 5.4/5.7.

Nick

[John]

The only other possibilities that occur to me are Deep Sky or Dark Mediterranean Blue, but I think they would be darker than 157.

Joseph

Hi Joseph. Your post prompted me to go back and look at the Deep Sky and DMB chips in the RAFM book. Interestingly, Deep Sky is very close indeed to BS381c-104 Azure. I don't know if 104 Azure has changed that much since it was 4 Azure in the 1930s but assuming the historic and current colours have at least a nodding acquaintance, we may have an idea where Deep Sky originates from. Deep Sky is much more "blue" than my version of 157, which is quite grey. DMB is darker still, I would place it about half way between 104 Azure and 105 Oxford Blue.

Steve and Graham:

I'm not sure how much 157 varied between the end of the Authentics and the integration of those colours into the standard range, and unless anyone has a tin of HB13 they can let us experiment on, I'm not sure we can take that line much further. Having said that, and acknowledging the memory is a fickle thing, I do remember buying the Authentic set for RAF Overseas colours shortly after the Airfix 1/72 Spitfire Vb came out. That was a big thing, because I couldn't get Authentics locally so a trip to either Glasgow or Edinburgh would have been needed. I painted the Spitfire in desert colours and can remember being surprised at the lack of contrast between the upper colours and how dark the undersurface was. Not scientific but it does suggest HB13 was a bit on the dark side in the late 1970s. I've still got the Airframe Silver from that set.

I don't have a copy of the Humbrol Colour System folder but if anyone has the Azure Blue formula from that, I'll give it a bash.

Mike:

I have the LifeColor paint somewhere, maybe at work. I'll dig it out and do a comparison. I'm also going to order some White Ensign colours to see how they compare. I should have an up-to-date Hornby tin of 157 lying around too.

John

[John]

I am surprised by the appearance of the Xtracolor sample as I had always thought - without actually checking - that Xtracolor's Azure Blue was one of the closest to the original. I used it on a Kittyhawk and found that UV exposure (of the model) caused a colour shift to a much "cooler", greener blue. This has resulted in the model now having a patchwork appearance underneath! On the face of it the photo of the sample doesn't look like it though. Mine is a very old tin - so has the Xtracolor Azure Blue also changed over time?

The FS 595B comparisons are interesting too, because although 35231 is cited as being closest (also by Geoff Thomas) in pure colour science terms 35240 is actually closer. Neither can be considered close matches though. In Munsell Azure Blue is another PB - Purple Blue - with a red pigment to it, actually 4.1 PB 6.2/6.0 so your perception of a "violet caste" is spot on. Geoff Thomas wrote that it was 5 PB 6.5/6.0. The difference between the two is 2.66 where <2.0 = a close match, with Geoff's being paler, perhaps the result of his measuring an actual faded sample.

To complete the math 35240 is 3.6 PB 5.6/5.5 and 35231 is 4.8 PB 5.4/5.7.

Nick

Hi Nick.

I was a bit surprised by the Xtracolor results too, although I do remember commenting in the past that their Azure Blue looked too bright on a model. My tin will be fairly old as well, at least 6 years, so the current crop could well be different. The Xtracrylic would be a more recent purchase, but even then it's a few years old now.

John

[Edgar]

In "The Colour System," Azure Blue is 49 of 34, 10 of 25, 1 of 33.

When the Authentics range died, we were told, at the time, that the cause was due to the Humbrol foreman matching each batch to a colour sample, taken from the previous batch, rather than the original master colour swatches, hence the gradual drift.

There was also a story that someone misread the High Altitude Fighter scheme as having Azure Blue undersides, and matching to that colour, when it was, in fact, PRU blue.

Edgar

[John]

In "The Colour System," Azure Blue is 49 of 34, 10 of 25, 1 of 33.

When the Authentics range died, we were told, at the time, that the cause was due to the Humbrol foreman matching each batch to a colour sample, taken from the previous batch, rather than the original master colour swatches, hence the gradual drift.

There was also a story that someone misread the High Altitude Fighter scheme as having Azure Blue undersides, and matching to that colour, when it was, in fact, PRU blue.

Edgar

Thanks Edgar. I'll convert these numbers into percentages and see what we get.

John

[Bo hermansen]

Hallo John

I would look up White Ensign models Azure blue, WEM have some great colors and they always keep

their colour up-to-date, they changes lots of their RLM colors when Merricks book came out.

I tried using it on my guniea pig BF 109, and it look close to the color chip on uor RAF chip card

Bo

I

I'll track down some other attempts at AB and see how they compare, both with each other and with the sample chip.

John

[stevehnz]

Thanks Edgar. I'll convert these numbers into percentages and see what we get.

John

In my example above, sorry, not too clear or well lit, I used the mix that Edgar has given but called it 1part of 25, to 5 parts of 34 to 0.1parts of 33 for which I used a tiny dab on the end of a brush to go with the drops from my pipette. If I was mixing enough to do a model I would do the whole mix. I felt the tiny dab of 33 I got might have been slightly to much hence I got a slightly greyer effect than I wanted. Better to add too little to begin with & add more if need be. I've found a source of Aeromaster azure blue & put some aside as I think theirs is about right. It compares quite well with the Humbrol mix.

Steve

[Nick Millman]

I found that I had a tin of HB13 Azure Blue with the black and gold label so I tested it today together with Xtracolor X26 and Aeromaster 9116. It is 1.7 PB 4.2/1.6 and compared to the RAF Museum chip the difference calculation is a massive 22.7 where < 2.0 = a close match. It is very close to FS 595B 36118 @ 2.85! A rather dark blue-grey. Unfortunately I don't have any 157 to compare it with.

Xtracolor is 1.4 PB 5.1/7.0 @ 11.1 and Aeromaster is 2.1 PB 5.9/4.7 @ 4.59. The Xtracolor surprised me because it looks fine until actually measured and compared. My sample is not far from the appearance of John's.

Nick

[John]

I found that I had a tin of HB13 Azure Blue with the black and gold label so I tested it today together with Xtracolor X26 and Aeromaster 9116. It is 1.7 PB 4.2/1.6 and compared to the RAF Museum chip the difference calculation is a massive 22.7 where < 2.0 = a close match. It is very close to FS 595B 36118 @ 2.85! A rather dark blue-grey. Unfortunately I don't have any 157 to compare it with.

Xtracolor is 1.4 PB 5.1/7.0 @ 11.1 and Aeromaster is 2.1 PB 5.9/4.7 @ 4.59. The Xtracolor surprised me because it looks fine until actually measured and compared. My sample is not far from the appearance of John's.

Nick

Very interesting indeed! I was just looking at the swatch I brushed out and thinking it would be a reasonable approximation of Extra Dark Sea Grey.

I'm beginning to wonder if Xtracolor and so on are matched to the approximate FS595 value, which looks to be a purer blue than AB. Looking at the mix described by Edgar and Steve, I think it may have the same problem in that it doesn't have any red in it to give the slightly violet appearance that is apparent in fresh AB.

John

[tangerine_sedge]

This is one of those rare occasions when being colour blind is a bonus. Most of those Blues look very similar to me (If I really concentrate I can see some differences, but at a casual glance they are the same). In this case Ignorance is bliss for me

[John]

I was in town this afternoon so, in the interests of research, I went to the model shop and bought a tin of Hornby vintage 157 to see what their current take on Azure Blue is. Having brushed it out I can say that it's pretty much identical to the old Humbrol colour I tried earlier. It looks like Hornby's current supplier did a very good job of matching the old Humbrol paint, which is presumably what they were asked to do.

John

[Hardtarget]

Interesting discussion here, and i too have been looking for a replacement to my Gloy colour azure blue. To me this is the most accurate to that chart as can be seen here on my recent hurricane for the matchbox GB.

This was taken outside under shade so the blue looks darker than in the flesh. If you have hu157 i belive adding some white until you get the match.

[stevehnz]

Very interesting indeed! I was just looking at the swatch I brushed out and thinking it would be a reasonable approximation of Extra Dark Sea Grey.

I'm beginning to wonder if Xtracolor and so on are matched to the approximate FS595 value, which looks to be a purer blue than AB. Looking at the mix described by Edgar and Steve, I think it may have the same problem in that it doesn't have any red in it to give the slightly violet appearance that is apparent in fresh AB.

John

John, have a look on THIS page which is where I got the above mix from that Edgar mentioned. Look in the US Airforce,US navy WWII group at the Azure mix for that, it has red in it & samples I have seen to that standard appear to be more violet that the RAF colour. I've no way of knowing how correct this is but dare I say that colour pics such as I have seen appear to give a more washed out smokey effect rather than violet. As always such a subjective topic & when weathering & scale effect etc come into it I think I would be best to duck for cover. Reading Nicks replies above as well, is it fair to assume that we actually have two colours called Azure blue, one Britiah & one from the USA.? I'm wondering which one the Munsell value that Nick refers to with the red in applies to.

Steve.

[Nick Millman]

John, have a look on THIS page which is where I got the above mix from that Edgar mentioned. Look in the US Airforce,US navy WWII group at the Azure mix for that, it has red in it & samples I have seen to that standard appear to be more violet that the RAF colour. I've no way of knowing how correct this is but dare I say that colour pics such as I have seen appear to give a more washed out smokey effect rather than violet. As always such a subjective topic & when weathering & scale effect etc come into it I think I would be best to duck for cover. Reading Nicks replies above as well, is it fair to assume that we actually have two colours called Azure blue, one Britiah & one from the USA.? I'm wondering which one the Munsell value that Nick refers to with the red in applies to.

Steve.

Hi Steve, mine was measured from the RAF Museum chip but, yes, they are slightly different. I have a US ANA 609 chip too but I haven't measured it because it is slightly cracked and chalked. It is not the same colour but what you describe about their characteristics is in reverse. The US chip is a slightly deeper, cooler blue. The RAF colour is bright and clear with a "warm", slightly violet caste.

I'll have a go with the corn oil tomorrow and see if I can get a measurement off the ANA 609 to compare.

Treat all FS equivalents with a pinch of salt unless they come with a difference calculation! I could write a mini series on what happens when we attempt to match visually but put simply many visual matches between two standards or paint chips tend to result in one of two things -

1. The colour that is closer in relative brightness but not so close in hue is selected; or

2. The closest hue is selected and there is less sensitivity to the brightness differences.

I call this the "White Rabbit Theory" in memory of and in tribute to a much respected and greatly missed colour scientist. People can get quite indignant about it ("I know what I'm seeing!" etc.) but they don't even realise it happens. This is one of the reasons I don't trust visual matches (there are other reasons too) - and no offence to anyone is intended or should be taken - but prefer instead to number crunch the tristimulus values, the chromacity coordinates and the CIE LAB values!

Nick

[Kes]

I have Authentic HB13 and a modern 'Super Enamel Hu157 which I could paint some swatches for comparison if this could be of some use? For what it's worth, the HB13 seems greyer and Hu157 seems brighter if the tin lid colours are anything to go by?

Anyone know how Humbrol researched and arrived at the colours they produced in their Authentic Range?

I don't doubt errors were made, their EDSG HX6 is decidedly green? and the Dk Greens, WWII HB1 and modern HX1 are quite different? should they be the same with a matt and semi gloss finish? Still closer than the recomended Hu30!! Sorry for the digression from the original subject, maybe I should open a new thread!!

. . . Kes

[stevehnz]

I call this the "White Rabbit Theory" in memory of and in tribute to a much respected and greatly missed colour scientist. People can get quite indignant about it ("I know what I'm seeing!" etc.) but they don't even realise it happens. This is one of the reasons I don't trust visual matches (there are other reasons too) - and no offence to anyone is intended or should be taken - but prefer instead to number crunch the tristimulus values, the chromacity coordinates and the CIE LAB values!

Nick

I guess that shoots my TLAR system of colour matching right out of the water Nick I think I get what you're on about, I have used THIS site to look for matches & while not pretending to understand it it has shown me something of the complexity of colour composition & that two paints that look very similar can be relatively different in their make up. I don't spose you have a link to a "Munsell for numpties" or something similar, it'd be nice to know more.

Just so this isn't totally off topic, I'll put some colour patches on the underside of the Vindicator wing so as to compare it with the H157 & Aeromaster 9116, which to me is what I want from an Azure Blue per the TLAR system. I might also try the Humbrol colour system mix with a teeny spot of red in it.

Steve.

[John]

I have Authentic HB13 and a modern 'Super Enamel Hu157 which I could paint some swatches for comparison if this could be of some use? For what it's worth, the HB13 seems greyer and Hu157 seems brighter if the tin lid colours are anything to go by?

Anyone know how Humbrol researched and arrived at the colours they produced in their Authentic Range?

I don't doubt errors were made, their EDSG HX6 is decidedly green? and the Dk Greens, WWII HB1 and modern HX1 are quite different? should they be the same with a matt and semi gloss finish? Still closer than the recomended Hu30!! Sorry for the digression from the original subject, maybe I should open a new thread!!

. . . Kes

I'd be wary of using Humbrol tin lids as an indicator of anything! The paint in the current 157 tin is markedly different from the colour on the lid!

John

[Nick Millman]

I guess that shoots my TLAR system of colour matching right out of the water Nick I think I get what you're on about, I have used THIS site to look for matches & while not pretending to understand it it has shown me something of the complexity of colour composition & that two paints that look very similar can be relatively different in their make up. I don't spose you have a link to a "Munsell for numpties" or something similar, it'd be nice to know more.

Just so this isn't totally off topic, I'll put some colour patches on the underside of the Vindicator wing so as to compare it with the H157 & Aeromaster 9116, which to me is what I want from an Azure Blue per the TLAR system. I might also try the Humbrol colour system mix with a teeny spot of red in it.

Steve.

Hi Steve

That's an interesting site with a neat tool but the translation of paint surfaces to digital colour values - and vice versa - is complex. You can take a digital reading from an image but in reality the pixelation of digital images limits the certainty of the values and creates a wide margin of error.

The best place to start is probably the Wiki entry on Munsell as it provides a basic overview and has links to explore further:-

Munsell Color System

Beyond that the Munsell Color Science Laboratory has lots of interesting stuff too:-

Munsell Color Science Laboratory

One of the main problems is that the Munsell color system is just not as accessible as, say, FS 595B, and the values are meaningless to many modellers (some get quite annoyed about them). Methuen ditto. Then the reliance on FS595B creates a situation where in the absence of any other comparisons the closest FS (often not that close at all) is sometimes presumed to be the colour. Again, depending on the actual colour this may or may not be much of an issue. At the end of the day colour precision has a limited place in modelling which is mainly interpretive. More important to modellers is to get the character of a colour right - or convincing - but exploring the maths can help to identify when colours are way off and to pin down their essential character. There are millions of colours but not so many pigments and they work together in predictable ways, so underlying the multiple variations are much fewer groupings which may be categorised. Many modellers tend to focus on the surface appearance without trying to understand how the pigments work (that is a simplification of course) and there are a few curved balls and surprises in this that mean that, without knowing them, it can be very hard to recreate the essential appearance of a colour.

The other aspect that gets confused is the difference between paint colour standards and the supplied paint. In this thread there is a classic comparison between the two, where samples of paint are compared directly to the standard, but I think it is all too easy to presume that the standard is the paint colour and to forget the inevitable variations in the supply of real paint to match it. By this I don't mean anything goes - but it leads back to the essential character of the intended paint colour. There is an inevitable interpretative skill that comes into play for modellers, because they have to take into account not just this, but other factors like scale, weathering, etc. So the two approaches, colour precision pedantry and artistic interpretation are not mutually exclusive. The one can inform the other without becoming a burden to either. The more I look at it the more I realise variations don't matter - in fact are impossible to eliminate because they are almost infinite.

I'm sorry if this is a bit of a digression from the topic in hand!

Nick

[John]

Here's what I got when I tried the mix mentioned by Edgar and Steve:

The result is lighter and brighter than either of the Hannants shades, so heading in the right direction. Compared to the RAFM chip (and acknowledging Nick's points above!) it still looks too grey and not violet enough to me.

This is obviously an elusive shade. I'm wondering if Humbrol white plus 25 Blue is the correct way to start, I'm not sure 25 is strong enough.

John