"True" Irish railway liveries

[Markleman]

I have been reading another thread about various shades of Green livery used on CIÉ in days gone by.

When it comes to interpreting old colour images I have the following thoughts and I wonder what others think. They are all about how I see the colours. I bow to the greater knowledge of those who saw them in real life (if their colour vision has been tested).

1) The colour response of the film stock in use varied enormously. At different times I used Agfacolor CT18 and 21, Agfachrome, Kodachrome, Ektachrome, and Fujichrome. To this day I can work out which of these was individually used for almost all of my slides. The E6 process slides come the closest in colour but vary a lot in contrast, while the earlier processes like the Agfa and Kodak ones have quite different colour rendition. I also used GAF Ascochrome and ORWOchrome occasionally, and they were totally different again, GAF had low colour staturation and high contrast, ORWO was over-saturated to my eye. Two photos of the same train taken on different film stock look quite different (often on a trip I would be using Agfa or Fuji for 35mm slides and Ektachrome on 120 slides). Even back then, and nobody liked to think about it, we all knew that different batches of the same film varied a bit too. Variation in my cine films are even more noticeable.

What film people will have been using in the 1960s may well be unknown today. If they had the processing done unmounted or used bulk film, the slide mounts might not have a brand on them. Not much point trying to work this out from my slide mounts.  I have 35mm Fujifilm and Ektachrome slides in Agfa mounts. My 120 slides are in Gepe plastic or Widescreen Centre cardboard mounts. We did what we had to do in order to make this hobby affordable.

2) Back in the days when film processing was expensive lots of people took steps to cut costs. Some would buy bulk film (often unbranded old 35mm movie stock) and load it into old cassettes. Ferrania film was also sold for this purpose. In addition, although this next point should not in theory affect the colour balance, lots of people bought the chemicals and did their own E-6 processing to save money. Otherwise non-prepaid E-6 processing was entrusted to some photo lab via the local chemist. Net result for colour accuracy of unknown film and unknown processing - doubly unknown.

3) Over time dyes fade. Some more than others. This is where using variable film stock can show up. Some brands are reputed to fade in colour terms more quickly than others. Colour film has been in fairly widespread use since, say 1950, which is 75 years for the chemicals in the film to degrade. OK, they are "fixed" in the original processing, but I doubt if the makers reckoned we should expect perfection after that period of time. And sometimes one primary colour dye can fade more than another.

4) Anything scanned from a print is likely to be pretty far from the original colour balance (which isn't that good on prints anyway).

5) Many of us have inbuilt colour "abnormalities". At one time I work in a photofinishers in Belfast, making colour prints from colour negatives using manual colour balancing. This was a very skilled process and I was not very good at it. I was moved to the interneg line where at least I was using slides to make (mostly) slides and the prints I made could be colour balanced by somebody better than I was at that. What I did not know was that my problems were caused by me being "colour blind". Not that I see only in black and white, but I do not see reds and green quite the way I would like to. If I had not taken an Ishihara test I would never have known. Lots of us have slight variations in how we see colour. 1 in 12 males has congenital red/green colour blindness. Work out how many males are on this site, take a twelfth of that, and you have quite a few people. This is why they changed the colours in electrical cables years ago (train drivers have colour blindness tests). It can be quite slight and hard to spot. It does not only affect red and green, but all the other colours made up from those two primary colours.

6) Anything that has been scanned in will have had to undergo analysis by the software doing the processing as well as going through the optics of the scanner. Mass produced scanner LEDs may not be true colour temperature to start with and then may change over time. Ordinary desktop scanners are often not the greatest. The makers of the slide film have gone through years of painstaking research to try to make their colours true, and then somebody scans them in an unknown scanner. After that they often get processed using something like Photoshop though a lot of scanner software has "enhancement" built in. This involves having automatic filters applied or else they are subjected to colour filtering by the person who scanned them, with all their bias and potential colour blindness. Then we view them on our computer screens, many of which are not properly colour balanced. I have a SpyderPro display screen colour balancing kit (they aren't that expensive) but frankly, we are into guesswork here.

7) Beware copied slides. Just like making prints, you are introducing scope for more variation. Even when done on a professional interneg machine this is not a great process, but when done on the cheap slide copiers we used to have it is disastrous for colour accuracy.

So, the conclusion I have come to over the years is that my slides are only a broad representation of the colours that were there when I took the shot. The colours are never exact. My slides are all I have. Comparing colours within one slide is pretty accurate, but switching between slides of different origin, variable sun angle and cloud cover, different dates and with different fading on the paint on the real thing, becomes a bit of a nightmare. Big shifts in livery are easy to see, subtle differences in shade may be due to any or all of the factors above.

I think that most colour rendition is like the politicians we have. Untrustworthy, rather vague, inclined to be not accurate, but the only options we have. Unless we were to elect a dictator of course, but that could hardly happen, could it?

[Tullygrainey]

Many thanks for this. It's very sound evidence for what many of us probably suspected but couldn't have articulated so thoroughly. You've certainly helped me stop worrying whether my County Down locos are the right shade of green 

Alan

[Galteemore]

There’s a lot in this. I think the best we can generally hope for as modellers is an acceptable range of a particular colour. I have recently been experimenting with GN loco blue on a HD body, and there’s an inexact sweet spot which just looks right based on taking some kind of average of published pics! My own experiments are a work in progress.
 

Ian Rathbone, who is the guru of model loco painting, says in his book that GNRI blue is basically impossible to define! And he’s painted a few. Two of his paint jobs below. One of these looks too light, but is what his client supplied. The other looks more on the money. Fred Graham’s were painted with genuine Dundalk paint but actually look a if anything little dark. 

Edited Sunday at 15:22 by Galteemore

[Tullygrainey]

Is there a 'scale' element to colour as well as everything else? Colours that are accurate to the prototype but look wrong when applied onto a model. It's often the case that pure white looks wrong on a model and needs to be toned down with a bit of grey to look 'correct'. 

Those Rathbone paint jobs are exquisite. I picked up a second-hand copy of his book recently. It's really informative and totally demoralising in equal measure!

[Markleman]

i think that if it looks right to you, then it is right. Each of those blue shades looked right to somebody, but yet they all look different to me. Blues I can do.

My point is that once you take a photo of it, even a digital one, and post it to a site or publish it in a book, then it is likely to look slightly different. [Edit - if we work from the colours in site images and books, then the same applies]

There may be a scale element as the eye/brain are constantly making adjustments.

My layout is under flourescent lights and looks different in photos unless I correct the colour temperature in the camera settings. Daylight varies throughout the day, the eye constantly corrects for this, the camera doesn't. Thus different film types respond differently.

So maybe we should hang loose and accept there will be small variations in the way colours are reproduced in books and on computer screens. 

Anyway, it gives us plenty to talk about.

Jim

Edited Sunday at 18:20 by Markleman
D'oh

[David Holman]

A further point, which I have mentioned before, is that males of our species start to lose colour definition with age - a bit like the film. In this case it is appreciation of green that goes. Not good for the Irish scene! Read somewhere that we should check with the lady in our life as the female of the species does not suffer from colour matching issues as much - but am guessing most of us knew that anyway...

 As Jim says, light is probably the most important thing. The light when the photo was taken and the light the model is painted and viewed under. JHB of this parish has done a lot of work on identifying the paint 'code numbers' of many of the livery colours, presumably based on analysis of the real thing, where available, so this gives a bit of a bench mark.

 Equally, as soon as anything is painted, it quickly comes under the influence of weathering, in all its forms, from the accumulation of grime to fading from sunlight. Therefore, if our models are depicting a working scene (rather than intended for the show case), then weathering is key, which is why Martyn Walsh's book on the subject remains so important to me, though there are plenty of other options as well.

[Westcorkrailway]

Worse still. If you didn’t grow up with these things and are used to seeing the photos it can be jarring to see what the actual colour is. Multiple shades lighter or darker then your own head cannon 

Edited Monday at 06:36 by Westcorkrailway

[GSR 800]

One of the first books on Irish railways I read as a young lad gave me a solid impression that the Woolwich moguls and the turfburner were painted in a pale blue!

Copyright 1995 Tom Ferris

 

[jhb171achill]

I’ve been reading this with great interest, some very valid and interesting points being made.

Having a lifelong interest in liveries and how weathering affects them, I’ll post in more detail in a while. I’m up to me eyes for the next three weeks but I’ll keep an eye here in the meantime….

[murrayec]

Same here, my impression from this image was pale, pale blue!!

[jhb171achill]

24 minutes ago, murrayec said:

Same here, my impression from this image was pale, pale blue!!

And there we have it; photos need considerable interpretation.

The “G” is of course silver, the Woolwich 2.6.0 very badly work CIE (dark!) locomotive green, same as carriages, and the turf burner started off in standard Inchicore dark locomotive grey, but ended its extremely short life in CIE lighter green as applied to diesels!

[jhb171achill]

19 hours ago, Galteemore said:

There’s a lot in this. I think the best we can generally hope for as modellers is an acceptable range of a particular colour. I have recently been experimenting with GN loco blue on a HD body, and there’s an inexact sweet spot which just looks right based on taking some kind of average of published pics! My own experiments are a work in progress.
 

Ian Rathbone, who is the guru of model loco painting, says in his book that GNRI blue is basically impossible to define! And he’s painted a few. Two of his paint jobs below. One of these looks too light, but is what his client supplied. The other looks more on the money. Fred Graham’s were painted with genuine Dundalk paint but actually look a if anything little dark. 

I would have to respectfully disagree with Ian. No colour is impossible to reproduce if research is properly conducted.

Fred Graham’s blue is 100% spot on. The other is too light, and that’s that. 

A tiresome narrative was often propagated by some in the past that “sure, the GNR just went to a local paint shop and got whatever blue they had”; the implication being that the blue varied considerably. This narrative seems, as far as I’m aware, to have originated at Whitehead when 171 had its first repaint by the RPSI - which was way, way, way too light; in order to justify same.

Such a tale, apart from being utter, abject nonsense, is an insult to those who designed and maintained the very carefully nurtured corporate images of railway companies.

Each railway company (bar the smallest) employed people in their works, whose job it was to maintain “recipes” for paints and pigments to ensure uniformity and continuity. The last such at Dundalk was Marcus Bailie-Gage, whom I knew well; he was a very close family friend going back to when my grandfather was with the GSR in Inchicore.

Marcus went to the great big paint shop in the sky about twenty years ago, and though an extremely mild-mannered man, would have spat nails at anyone suggesting the GNR just used any oul blue paint!

Now, of course, blue is a primary colour and is notoriously hard to keep fresh, and daily coatings of coal dust, oil and oily rags didn’t help. But that’s a different issue; GNR blue was uniform.

The same, incidentally, applies to the two shades of CIE green,  plus the oddly-titled “eau-de-nil” (“water of nothing”?) which they used for lining. The same narrative, “ah, sure, there were loads of greens” is again 100% completely wrong.

There were two.

1. Darker: Buses, lorries, station paintwork, platform wheelbarrows, coaches up to 1945, D class diesels when new, and steam engines which received green.

2. Lighter: Coaches 1955-62, and diesels repainted from silver circa 1957/8-1962.

But - especially the darker shade - it was applied to metal and timber surfaces, with varying degrees of exposure to weather, steam, hot oil, coal dust and filthy diesel fumes. Moreover, and perhaps far more importantly, it wasn’t renewed too often, as money was tight.

Thus, with GNR blue, CIE green, AND other liveries, there never were “multiple” variations of the livery, but there certainly were multiple variations of fading, weathering, and erosion of primary pigments. On CIE, many older vehicles in green were being photographed in colour for the first time when they hadn’t actually seen a touch-up paintbrush for a decade! 

I’ll expound on this more when I get a minute. The issues around perception, especially (as mentioned above) amongst ageing male persons, is a wholly different matter.

 

 

 

Edited Monday at 11:16 by jhb171achill

[jhb171achill]

Footnote to the above; I have seen a photo of a black’n’tan coach which LOOKS navy blue and tan; and all manner of other “photographic evidence” which - isn’t!

The late Cyril Fry took a large number of photos. Most were black and white (his earliest ones are around 1925-30). But for colour, he unfortunately used the cheapest and nastiest film he could get; thus his slides are in many cases useless. One, which would be a beauty if the colours had not decomposed, shows a “brown” B101 in a cutting on the Tullow branch. The grass and gorse on the sides of the cutting are purple…..

Edited Monday at 11:26 by jhb171achill

[Tullygrainey]

2 hours ago, murrayec said:

Same here, my impression from this image was pale, pale blue!!

That one threw me too, causing me to paint mine blue.

 

Subsequently corrected 

[Markleman]

1 hour ago, jhb171achill said:

The late Cyril Fry took a large number of photos. Most were black and white (his earliest ones are around 1925-30). But for colour, he unfortunately used the cheapest and nastiest film he could get; thus his slides are in many cases useless. One, which would be a beauty if the colours had not decomposed, shows a “brown” B101 in a cutting on the Tullow branch. The grass and gorse on the sides of the cutting are purple….

Excellent point.

These days those colours could be "corrected" by the touch of a button using Photoshop et al. That would create an impression of what the software designers thought they should be like. This would be fairly nearly correct for most purposes, but maybe not for copying the livery for my most precious models. Then human intervention could tweak it to a really good version of what (say) I thought it should be. You might have a different view on what I had done. But the slides then would not be "useless". This is worth considering in Fry's case. I bet it was Ferrania film he used.

Even the best film can deteriorate (don't keep them in the loft!). I wonder can I find an example to post here ...

[Markleman]

Here is one (if I can upload it). I fiddled with it as a possibility for Volume 1 of Line by Line, but it was never going to get into the book. The top photo is as it was (subject to the unaided scanning process) and the bottom one is how it ended up before I gave up on it. I could have made it better but I abandoned it. It was taken on 13 July 1974 on Kodachrome 64 and had plenty of time to fade.

Do you think that the upper or lower photo better represents the colour of the gold on the loco? Or red on the buffer beam? Good questions. The faded photo is almost certainly not right as you can see the colour cast on everything. The "enhanced" version on the bottom is not right either because it is what Photoshop thinks it should be. Putting colour images through Photoshop or similar things does not recover the original colours, it just changes them to what the software thinks it should be. OK , that is often a pretty good version, but it will never be the same as the scanned photo, and that photo will never be what the colour of the loco was thanks to colour balance issues with the film and dye fading over the years.

Yep, old photos great for identifying the broad colour, not so good for the fine gradations of hue.

[murrayec]

This is how I would push/adjust that photo. The photo is low resolution, better results can be achieved with higher resolution files.

Eoin

[Galteemore]

Looks really good. Totally off topic but that’s a bare 17 years after the SLNC closed -1974  seems a lifetime ago now! 

[Markleman]

I manipulated that photo, now it has been manipulated again. It looks better each time, but is it more accurate?

It is worth considering that many of the images we find on sites and in books have been manipulated.

True colours? Don't bank on it.

When I took it there was no yellow colour cast but my memory is now a bit hazy, like the photo.

[Broithe]

There are no true colours, particularly in photographs - just look at something which is half in sunlight and half in shade.

[jhb171achill]

1 hour ago, murrayec said:

This is how I would push/adjust that photo. The photo is low resolution, better results can be achieved with higher resolution files.

Eoin

Yes, exactly. 

We can only guess at what some early liveries were like, but in most cases the best option is a photo like the above, adjusted as best it can be, WITH a wide consensus of verification from those who SAW the thing in real life.

To me, that particular adjusted photo is 100% accurate.

[dropshort105]

A lot of Cyril Fry's models were liveried with the correct paint which he got from Inchicore, but as somebody said earlier, correct colour may not scale down very well, I go with " paint what you see " but that doesn't work with b&w or poor quality photos.

[jhb171achill]

23 minutes ago, dropshort105 said:

A lot of Cyril Fry's models were liveried with the correct paint which he got from Inchicore, but as somebody said earlier, correct colour may not scale down very well, I go with " paint what you see " but that doesn't work with b&w or poor quality photos.

Indeed, and other places too (Dundalk, York Road, Aughnacloy, Stranorlar and Queen's Quay).

A small number of his models are inaccurate, usually by choice (2 of his locos which never anything but grey are in GSR 800 class green, because he liked it!), it because in wartime he couldn’t get the right paint (his GNR “S” class).

But there are one or two surprisingly basic errors like his Lough Swilly tank engine, which bears a shade of green light years too light.

Edited Monday at 23:41 by jhb171achill

[Mayner]

I friend once spoke about building a monochromatic North Wales narrow gauge diorama as the historic photos were mainly black and white! Gets over the argument over prototypical shades, but potentially challenging to execute.

[David Holman]

There are a couple of layouts recently that have done the monochrome thing and it can be quite effective. Have considered it myself, though the biggest problem is scenic scatter and fibre materials, which would need recolouring or painting as there appears to be no grey shades available.

[Colin_McLeod]

Colour is difficult to get right.  I thought the blue of the IRM Mk2b NIR blue and grey coaches was too dark.  Upon investigation I realise that it perfectly matches the MIR 'NIR light blue' paint which was professionally colour matched to actual NIR paint in the early 1980s.  

Despite this it still looks darker than my recollection.  Photographs don't help as there you will find a whole range of blue shades.

For my own interest I repainted two IRM coaches. One with a lighter blue stripe, and the other in NIR maroon and dark blue.

This is the result.

 

I think the repaint is closer to my memory but maybe a bit light, so further experimentation will follow.

The MM GM also looks on the light side and I may repaint that if I can get the shade right.

[jhb171achill]

13 hours ago, Colin_McLeod said:

Colour is difficult to get right.  I thought the blue of the IRM Mk2b NIR blue and grey coaches was too dark.  Upon investigation I realise that it perfectly matches the MIR 'NIR light blue' paint which was professionally colour matched to actual NIR paint in the early 1980s.  

Despite this it still looks darker than my recollection.  Photographs don't help as there you will find a whole range of blue shades.

For my own interest I repainted two IRM coaches. One with a lighter blue stripe, and the other in NIR maroon and dark blue.

This is the result.

 

I think the repaint is closer to my memory but maybe a bit light, so further experimentation will follow.

The MM GM also looks on the light side and I may repaint that if I can get the shade right.

I would say the left hand and middle coach are spot on - right hand one I’d say the blue is a BIT dark.

[David Holman]

We were talking about colours down the Club last night, though in this case it was roof slates, where there are certainly a lot more than 50 shades of grey.

 Ultimately, you can spend any amount of time trying to get something 'right'. However, this also means doing the same thing on everything else. Fine if it is a showcase model, weathered or pristine, but from a whole layout point of view, it starts to get difficult. In any case, a single 'perfect' wagon or coach is arguably unnecessary from a whole train point of view - especially if the train is moving. It is the overall impression that counts. Likewise buildings and scenery, where the eye registers the general scene. One stunning model among stuff of lower standards will stick out in the same way as a poor one.

 So, consistency may be more important that total accuracy, perhaps?

 

[Westcorkrailway]

2 hours ago, David Holman said:

 

 So, consistency may be more important that total accuracy, perhaps?

 

I subscribe to this when making my models, problem is how impossible it is to pull off! RTR manufacturers will naturally have slightly different shades. Look at A46 vs the hattons genesis. Only a slight difference but the same livery. 
 

I think it would be worth IRM’s or any other manufacturers time selling the paint that could In theory match its stock if you use primer, even rattle cans of it. Though of course this might not be feasible

 

also, one must consider there was a bit on inconsistency in reality too. Whole can of worms  

[Horsetan]

1 hour ago, Westcorkrailway said:

....I think it would be worth IRM’s or any other manufacturers time selling the paint that could In theory match its stock if you use primer, even rattle cans of it. Though of course this might not be feasible...  

You'd need to find out who supplies the factory in China with its paints. Good luck with that. 

[minister_for_hardship]

Don't forget, different batches of what are supposed to be the same shade of painted product will appear dissimilar when placed side by side. Learnt that in a previous life in the construction industry! It is the case in real life and models.

[Garfield]

2 hours ago, Horsetan said:

You'd need to find out who supplies the factory in China with its paints. Good luck with that. 

We know quite well who supplies the paint to our factories.  

Paint shades are based on colour matching systems (e.g. RAL, Pantone) rather than a range offered by a specific manufacturer.

We previously looked at the whole area of selling paint and it's an area that requires a significant amount of 'heavy lifting' to make possible, and the Irish market is too small to justify minimum production quantities at this time.