Colourisation: problems ahead.

[BachelorBoy]

Black-and-white photos are colourised frequently these days.

 

I am worried (not just for railway modellers) that this is going to make historical research more difficult. 

 

Colourisation is becoming much better. But if buses, or shopfronts, or railway buildings and rolling stock are included, the colours are only going to be guesswork. Plausible, but still guess work. 

 

Once colourised digital photos start circulating online, detached from the original post that may say they've been colourised, it's going to make it very difficult to tell which photos are real, and what are colourised.

 

This is not going to help.

 

 

[Wheatley]

4 hours ago, BachelorBoy said:

 

I am worried (not just for railway modellers) that this is going to make historical research more difficult. 

For hobby purposes perhaps, but any sort of academic or quasi-academic research will still require reference to primary or at least credible sources, not a colourised jpeg off a Facebook local history group. 

 

Although given hobbyists' collective tendency to start fights in empty rooms where we do know what the real colour is, I doubt this will make much difference. 

Edited March 12 by Wheatley

[Mark Forrest]

Old colour photos arguably aren't much better as colour rendition might not have been great to start with and prints/slides fade over time.  Probably best to treat any colour image as an approximation.

[Bernard Lamb]

A coloured print from a linocut, after an original drawing by John Mennie. Thailand May 1943.

 

 

 

 

A photograph by me, November 2018. Not far from the location of the original drawing . It was difficult to find the exact spot.

 

 

Colour certainly can vary.

Bernard

[LBRJ]

There is also the point that everyone pretty much sees colours differently anyway, at least to some degree.

[SRman]

Back in the 1980s I took a good many photos of my travels in England and Scotland, and had most of the films developed and printed in a shop in Leeds (Cross Gates). The last two films had to come back to Australia with me and were put into a local shop for printing. The colour renditions were quite different on the locally printed ones.

So, we can add the inks sued in the printing process to the potential variations.

Then again, when looking online, we can have different monitors with different colour renditions, and we can personalise them with different colour temperatures, brightnesses and contrasts. I found that my top quality Dell Ultrasharp monitors at my work were much better for colour matching than my good quality Asus home monitor, and that was still better than the laptop screen. At one time I was assisting a local model manufacturer with getting colours right on an Australian bus model they were producing, and the only monitor I could use was my work one: matching yellows and creams was particularly difficult on all the other screens.

In short, the only real source for accurate colour matches for our models is the prototype, assuming they haven't faded or darkened. There were also various livery books available that had colour swatches matched to paint flakes from prototypes - samples that are sometimes buried under layers of paint from subsequent liveries.

Is near enough good enough. Being so far from Britain as I am, near enough often has to suffice for my own modelling attempts.

[Michael Hodgson]

Not a problem.  We all know that Stroudley painted LBSCR engines an improved shade of green.

[SRman]

15 minutes ago, Michael Hodgson said:

Not a problem.  We all know that Stroudley painted LBSCR engines an improved shade of green.

 

If he said it's green, then it's green! 😉😂

[kevinlms]

42 minutes ago, SRman said:

 

If he said it's green, then it's green! 😉😂

But that's mainly because IIRC, he resented anyone questioning his decision, so green it was!

[kevinlms]

1 hour ago, LBRJ said:

There is also the point that everyone pretty much sees colours differently anyway, at least to some degree.

Colour aside, think of the big debates that sometimes occurs, when people try to identify a location, or the type of loco/rolling stock, of a photo.

 

Books are notorious too, on how they print a photo, sometimes top quality reproductions (such as Wild Swan etc) and very grainy, such as some Ian Allan books and plenty of other publishers.

 

So colour is sometimes the least of the problems!

[Steamport Southport]

1 hour ago, Michael Hodgson said:

Not a problem.  We all know that Stroudley painted LBSCR engines an improved shade of green.

 

All down to a misquote.

 

What was actually said is it's "an improvement on Engine Green" which was the livery of HR locomotives at the time. He took the livery to the LBSCR.

 

Any idea that he was colour blind is an urban myth I'm afraid.

 

 

Jason

[Hroth]

I tend not to criticise grainy photos as at the time of publication, the only existing photos of a rare something or other may well have been taken in poor conditions by an enthusiast with what effectivly amounts to a Box Brownie. I'd rather have a contemporary picture than not, even if the quality is relatively poor.

 

As for colour, early film aimed at the amateur market was expensive and rarely used. Unless in the cinematographic industry, professionals were unlikely to use colour to record railway subjects.  In addition, pre-WW2 colour film was very grainy and impressionistic in terms of colour balance, so a colour image prior to the late 50s may be helpful, but should be taken with a pinch of salt!

 

[bécasse]

When the use of colour film by individuals was still rare the processors could have an influence on how colour was rendered in the resulting slides. Our next door neighbour's elder daughter got married in the early 1950s and, as her newly betrothed was comfortably off, they went to southern Ireland for their honeymoon accompanied by a decent camera with colour film in it. When they got the slides back they were generally excellent but they were very surprised to find that the deliberately posed shot of the pair of them in front of an ex-GPO pillar box showed the box to be red. When they complained to Kodak, they received a very nice reply (together with a quantity of complimentary colour film) thanking them for the explanation that the box really had been green and not red. Apparently, it had been rendered correctly on the film but the individual processor noticed that what he thought was an obvious British post box hadn't come out the correct colour and so had "corrected" it. Kodak were indeed very pleased to learn that it wasn't a fault of the film but of the perceptions of the processor.

[Hroth]

18 hours ago, bécasse said:

When the use of colour film by individuals was still rare the processors could have an influence on how colour was rendered in the resulting slides. Our next door neighbour's elder daughter got married in the early 1950s and, as her newly betrothed was comfortably off, they went to southern Ireland for their honeymoon accompanied by a decent camera with colour film in it. When they got the slides back they were generally excellent but they were very surprised to find that the deliberately posed shot of the pair of them in front of an ex-GPO pillar box showed the box to be red. When they complained to Kodak, they received a very nice reply (together with a quantity of complimentary colour film) thanking them for the explanation that the box really had been green and not red. Apparently, it had been rendered correctly on the film but the individual processor noticed that what he thought was an obvious British post box hadn't come out the correct colour and so had "corrected" it. Kodak were indeed very pleased to learn that it wasn't a fault of the film but of the perceptions of the processor.

 

Yes, my parents went on honeymoon to Jersey in the early 50s, not quite so well off as my fathers camera, a Zeiss Ikon folder with a Tessar lens, only contained Dufaycolor slide film!  I still have the slides, but the Dufay process produced images that were rather grainy and less saturated than Kodachrome.

 

The other problem was that colour film of the time was relatively slow compared with monochrome film and light meters were not that common, even the extinction types were more guessing aids and a Weston meter was often as or more expensive than a medium range camera, so getting a correct exposure in conditions other than bright and cloudless was pot luck, so until colour film and metering was affordable, most amateurs stuck with mono.

 

And, as you say, colour balance was sometimes out of the photographers hands anyway!

 

[34theletterbetweenB&D]

On 12/03/2024 at 02:45, BachelorBoy said:

I am worried (not just for railway modellers) that this is going to make historical research more difficult. 

No need to be concerned. Any serious research into such starts from the knowledge that colour has no real independent existence, but is only an artefact of the illumination, reflection and absorbtion of light by the subject, and the response of our vision system. Stir in the effect of the gamuts of optics, film or image capture devices, processing and display and it makes the 

 

On 12/03/2024 at 06:50, Wheatley said:

...hobbyists' collective tendency to start fights in empty rooms...

over this matter positively hilarious.

[BachelorBoy]

11 hours ago, 34theletterbetweenB&D said:

No need to be concerned. Any serious research into such starts from the knowledge that colour has no real independent existence, but is only an artefact of the illumination, reflection and absorbtion of light by the subject, and the response of our vision system. Stir in the effect of the gamuts of optics, film or image capture devices, processing and display and it makes the 

 

Imagine you have a colourised picture of a BR Mark 1 you are trying to date or locate, or find out where particular stock worked at particular times. When colourised, it's maroon. But in fact the coach was Southern Region green. 

 

[Wheatley]

That's no different to inaccurate paintings or retouched photos. Or even trying to interpret shades of grey on a black and white photo (see parallel thread on the colour of wagon interiors). Sometimes you'll be wrong but all historical research comes down to informed educated guesswork, the degree of guessing reduces as the amount of reliable information increases. 

 

There are examples in aircraft modeling where people have followed published photos exactly, only for the uncensored version to appear years later out of the MOD or IWM vault showing that that particular Lancaster (or whatever) was festooned with aerials which the censor removed from the original print. 

 

At the end of the day, if it really matters (ie your academic reputation depends on it rather than  having to repaint a carriage or continue in blissful ignorance) then you'll either corroborate it with another reliable source or caveat it to warn that "this is based on xyz...". 

 

For further examples see the Victorian interpretation of an inguanadon with the thumbspike on its nose (its in a park somewhere) or the Horrible Histories sketch on paintings of medieval animals based on written descriptions. 

 

That's the joy of research. Sometimes it comes down to interpretation. 

Edited March 14 by Wheatley

[spamcan61]

4 hours ago, BachelorBoy said:

 

Imagine you have a colourised picture of a BR Mark 1 you are trying to date or locate, or find out where particular stock worked at particular times. When colourised, it's maroon. But in fact the coach was Southern Region green. 

 

The lining (or lack of( would be the key indicator there, irrespective of colourised or not 

[34theletterbetweenB&D]

7 hours ago, BachelorBoy said:

Imagine you have a colourised picture...

The correct approach starts right there. First task is to determine the origin of any image to assess whether it has been modified in any way. If it has, then it's for the bin in terms of all aspects of modification.

 

And there's much more: 'Preserved' items are a minefield. 'Locomotion' (Railway Museum) makes it quite clear that models derived from their collection specimens are models of 'as preserved'. Much as I enjoy my Stirling single model, I am fully aware that it represents a museum reconstruction of  how No1 looked at the end of its regular service. Bang on accurate for final GNR operating condition it cannot be.

[BachelorBoy]

4 hours ago, Wheatley said:

That's no different to inaccurate paintings or retouched photos.

 

It is very different, as it's so easy to do. Paintings take a lot of effort. Colourisation adds to the uncertainty.

 

[BachelorBoy]

3 hours ago, spamcan61 said:

The lining (or lack of( would be the key indicator there, irrespective of colourised or not 

 

Until someone who doesn't know it's been colourised says the photo proves there were maroon coaches without lining.

 

[34theletterbetweenB&D]

12 minutes ago, BachelorBoy said:

 

Until someone who doesn't know it's been colourised says the photo proves there were maroon coaches without lining.

 

No problem, we are entirely accustomed to informing those with insufficent information that they must revise their opinion.

[Mol_PMB]

1 hour ago, BachelorBoy said:

 

Until someone who doesn't know it's been colourised says the photo proves there were maroon coaches without lining.

 

There were, but the colour was instead described as 'crimson' ;-)

[melmerby]

On 12/03/2024 at 02:45, BachelorBoy said:

Once colourised digital photos start circulating online, detached from the original post that may say they've been colourised, it's going to make it very difficult to tell which photos are real, and what are colourised.

There's already colourised but inaccurate films around.

I posted in another forum about an easily researched but inaccurately rendered GWR train that had green & white carriages!😒

 

I have however seen some very well researched colourisations as well, such as the WW1 film "They Shall Not Grow Old" that film director Peter Jackson did.

[Wheatley]

The answer to all of these concerns is to cross check with other references if it's important enough to do so. There are enough non-colourised colour photos of Mk1s  to establish that within the relevant timeframe those on the Southern Region were mostly (?) green and those elsewhere were mostly maroon with a few brown and cream ones. 

 

Most of my reference photos for accurate train formations are black and white, and whilst they might be clear enough to determine which set of Comet sides I need, a lot of them are either too indistinct or too far away or too head on to pick out details like lining or shades of monochrome. Where I can see detail I follow it, elsewhere several of the RTR ones are the right coach type but whether they are maroon or crimson, line or unlined, depends on what was on release when I picked then up. Pragmatism, 80/20 Rule, bodging with a finescale edge, take your pick. Life is finite and the 'to do' list is long. 

 

I have to be precise at work. I don't necessarily at home.