Colour perception, especially with models, is an often debated topic especially when manufacturers occasionally, and some more than others, appear to get it wrong. There can be several reasons why colours on models do not always appear correct. In this post I look at some of the issues and reasons that can influence getting colours correct. I have been constructively critical in the past of some manufacturers attempts at getting colours / liveries correct and often try to get colours and liveries corrected, if possible, and have done so again only recently with some proposed SR locomotives (naming no names but fingers crossed they arrive OK).
I will use LSWR/SR Goods Wagon Brown as a case in point; for example, Hornby have had multiple attempts to achieve the correct SR colour. Back in 2016 their excellent SR Diagram 1530, as per my review here, was released in a good, if not very slightly too dark, representation of the SR Goods Wagon Brown.
Perhaps the fact that the official name of the correct dark brown colour is “Chocolate Brown” they chose milk chocolate instead?
Hornby have subsequent released further versions of this model in a darker version but is still slightly too light and lighter than the colour they used on the cattle truck!
In my day job I therefore, for my own satisfaction / reputation, had to ensure that the LSWR/SR Goods Wagon Brown on the Kernow Model Rail Centre ex LSWR Diagram 1541 Road Van that I was responsible for producing was as close to the correct colour as possible. I undertook a lot of research to be able to provide the factory with the correct paint references, although this is not as simple as it sounds as I will discuss below.
Following much historical research and checking many contemporary references I was able to provide the factory with a suitable British Standard paint colour reference, however even this is complicated by the fact that such a historical British Standard reference is now obsolete so not readily available for the factory to obtain! Careful checking and agreeing decorated samples ensured that I was happy with the factory’s interpretation of the colour to allow production to commence.
I am also only too happy to share my researched colour references with some other manufacturers, to try to achieve some consistency of colours for all Southern Railway modellers alike.
This included for example my good friends at Rapido Trains UK, and they specified with their factory my SR Goods Wagon Brown colour reference for their splendid ex SECR /SR open wagons.
I did however note in my review here, that their factory interpretation of the colour appeared lighter. I also thought at the time of writing that the finish of the model may have also affected the perception of the colour as it was a satin nearly glossy finish rather than matt.
To demonstrate this, I have now given one of the Rapido Trains UK wagons a simple single spray coat of Testers Dullcoat matt varnish and when pictured alongside the KMRC Road Van and one of the original Rapido Trains UK factory finished wagons the effect of the type finish and its perception of the same base colour can be clearly seen.
I will now apply the same treatment too all my Rapido Trains UK wagons from this batch (and I have also shared the results of this simple change of finish with Rapido Trains UK ).
It should be noted that I have purposely taken the comparison picture under the same lighting conditions. Different forms and types of lighting either when viewing the prototype, for example bright sunshine or a cloudy day, or models for example under warm or cool white lighting (see my post here about white is white…) can totally change the visual perception of a colour. I am also of course aware that you will be viewing this post on different devices and screens that will also create different perceptions of the colour!
In addition to historical superseded / obsolete colour references and paint finishes there are several other factors that need to be considered when specifying and choosing the correct colour.
Firstly, care should be taken when using old colour photographs, or for that matter preserved rolling stock, as there are so many variables that can affect the representation / comparison of any colour. As well as the lighting conditions at the time the image taken the use of different film stocks at the time and variations in any subsequent printing can give different colour hues. Something published as fact, even repeatedly or copied is not necessarily always factually correct and can still include errors or subjectivity.
Another factor to take into account especially with models is that of colour scaling; our perception of colour does not scale and will vary depending on the distance at which it is being viewed and also the size and the area of the colour, for example if you painted a model with exactly the same paint as a full-size example the model will appear darker when look at in isolation. This is therefore also an issue when using a small swatch of colour as an original reference, and this has been the case, in my opinion, with a small number of colours as referenced in otherwise excellent and well respected livery reference books.
Sometimes a model manufacturer will sometimes need to counter this by using a colour slightly lighter on the model than the full-size prototype so it ‘looks right’ to the eye.
It should also be noted that adjacent different colours to our chosen colour will affect the perception the hue, see the example shown left.
This is often highlighted when initially painting a model for example compare a lined and unlined model that uses the same base colour.
For example, a splendid malachite green Bulleid pacific will look to be a darker green until the three horizontal lines are added as can be seen in the image to the left of my 21c11 before and after lining has been applied and photographed under the same lighting conditions.
Finally, one further complication for model manufactures is the process used to recreate the often-complex liveries on a model. This is often achieved by a mixture of both paint and print applications, whereas the prototype is more often than not painted (although some modern liveries are via printed vinyls) . Different specifications are used for paint and print colours. For example, paint colours are usually specified to British Standard (both current and obsolete) or RAL numbers; whilst printing inks are usually referenced Pantone colours. There are often no direct conversions between some paint and print colours and errors can creep into the process when conversions take place. For example, sometimes a paint reference could give multiple close Pantone references, and it can even be the case that when some are converted back, they end up as a different RAL number!
It is therefore imperative that such conversions between paint and print references during the process are checked and agreed at every stage. It is the reason that creating an approved set of livery artworks must then be checked and further approved at the decorated sample stage (actual physical sample not photographs from the factory!) before production. Skipping some of these steps in the process, usually for apparent cost reasons, can easily result in mistakes, such as has occurred with the production of some models in the past and therefore be a false economy.
I hope this little walk through the hues and lows of the processes involved in getting the colour / more importantly, the perception of colours as correct as possible has been of interest, perhaps the first of an occasional “Insider insights” series? As always, I welcome and enjoy reading and responding to comments.
I work on the basis of holding ‘Graham Muzpratt’s opinion’ in high regard.
That way it appears to be very difficult to go wrong……
🙂
I now commit heresy quite frequently when painting buildings by getting the Phoenix Precision SR colours and mixing in a little matt light grey. The degree of “fading” looks better to my eye and more appropriate for a run-down 1950’s branch line. If the mixing is a little uneven so much the better!
Not heresy at all and always good to start with the correct starting colour as you have to create the faded look with.
Graham. Just to thank you for this excellent piece. I found it especially interesting since I have built and painted a number of LSWR wagons over the years. More recently I have purchased all of the RTR LSWR wagons you mention (avoiding the dreaded Hornby milk chocolate aberration!)
My take on this is that I am content to have small differences of tone between wagons ( and indeed coaches) but not drastic colour variations. Perhaps correct colour is like beauty in the eye of the beholder.
After 25 years working as a colour scientist, I have to say the only certainty is that it’s impossible to “get right”. There are probably something like 7 different types of “colour normal” vision, and depending on which group you fall into, you may see things differently from someone in one of the other groups. And it will vary then depending on the spectral content of the light under which the paint is viewed – you can have two different “warm white” LEDs with different spectral balances, so different people will make different colour matches under those apparently identical lights. Then there’s the problem that 1 in 7 men (and about 1-in 50 women) have colour deficient vision to a lesser or greater extent, at which point all bets are off.
Then of course you can have two shades of paint which look identical to the human eye under “Standard Illuminant D65” (average northern hemisphere daylight), but different under “Standard Illuminant C” (a tungsten bulb), because of varying reflectivity in the paint of different wavelengths which can make “white” match but other colours change relative to each other. And further, colours look different again to a camera, because a camera has different characteristics in terms of how it “sees colour” from the human eye.
The great get-out is that paint in the real world on the side of a loco, wagon or carriage will change over time due to sun/rain/cleaning/soot/brake dust, and paint composition, as well as the day of the week in the railway’s paint shop back in the days when paint was mixed on site to a “recipe”.
So the best we can hope for is making it “look roughly right”. And that’s in 12-inches to the foot preservation, without the “scale colour” effect which Graham refers to!
Very informative – thank you.
Hi Richard,
Thank you for adding the additional informative information.
Hi,
Interesting comments there. So if I’ve got 6 people in front of the layout all looking at the same thing they might not be seeing the same colour, if so might explain one or two comments. Plus might well explain the differences of opinion about one colour used on an early privstised unit. Might have to do some reading.
Thanks
Stu
P.S. I solved the Hornby SR-liveried LSWR Brake Van problem by masking out the lettering and thinly spraying over the brown with some black. It possibly now looks slightly too dark, but I’m happy (which is all that matters).
Hi Richard,
I completely repainted my Warner’s vans!
Phew! Thank you for the informative article. I had the same problems years ago with Southern EMU green on my layout Ashington. I have as many shades of green as EMU’s. Since building my kits, Hornby and Bachmann have produced models but even they have colour variations.
An insightful article. One other aspect which needs to be considered with photographs these days is that we look at them on a screen and the screen can vary in the colour regime it is working to. Ideally, the monitor screen should be calibrated. (How many people have done that I wonder.) And as we often use tablets & mobiles to view things I’m not even sure they can be calibrated. (And who has unwittingly tried editing images with the ‘nightlight’ on? ✋)
It’s a tricky area where old paint/pictures are concerned.
If we treat serious railway modelling as an art perhaps if it looks alright………..
Indeed and I did mention screen types along with different film stocks in passing within the article.
Hello Graham. A friend sends me your comments and observations from time to time, and this is one of those times. The “hues and lows” bring about some interesting discussions, and I wonder if anyone can ever truly replicate the original colour schemes. I am under the impression that much painting of rolling stock, certainly in the grouping era, was carried out in poorly lit sheds late at night where matching colours was a specific challenge. I also believee that paint production was still rather crude so that colour matching in both cases was a bit hit and miss. All of which begs the pedantic question of what is right and what is wrong? Just wondering.
Hi Roger,
Yes although Railway Companies they did have documented paint colour mixing specifications, and used reference colour swatches / plates, there would have been some variation between batches and works, but not to the extent of some of the errors on models we have seen. With colour scaling the effect would be actually less prominent in model form.
I am not sure where the impression “poorly lit sheds late at night” comes from as most primary livery painting was carried out at works and most had a dedicated paint shop area from quite early on that would have been working day shifts.
Hi Graham. Very interesting article. I hope Bachmann take note for their forthcoming Bullied coaches.
I can confirm that Bachmann have amended the the BRs green for the forthcoming new Bulleid coaches.
Reading all this again – it’s an interesting subject – and considering all the points made especially Richard Salmon’s additional comments it does seem that it would not be unreasonable to work on the basis that it is essentially impossible for the colour to be ‘right’ and that as long as it is ‘about right’ and it ‘looks acceptable to you’ it will be fine. Of course the degree of wrongness is infinite but for the LSWR/SR brown in question what we now have on more recent releases from different makers is within an acceptable tolerence of authenticity.
And if you disagree you could always paint your wagons any colour you like as long as you are happy. (I draw the line at purple!)