Experiments in monochrone

[57xx]

The topic of colours in old monochrome pictures often crops up in threads, most recently in the D299 Appreciation thread. On suggestion by @magmouse, I'm opening this thread so we do not to go too far off topic and clutter up the D299 one (or any other for that matter!)

 

A few years back I had a dabble in Photoshop using adjustment layers to try and replicate orthographic emulsions that were prevalent in the early 1900s rather than just hitting that "Convert to greyscale" button. I dabbled and then it gathered dust until recently when the topic of old monochrome pics and what originals colours in the pics may have or may not have been cropped up several times. So I dug out my old templates had a play and then realised - I could do better. So after several weeks of researching, I found the original info I had based my filters on was wrong. I found better (and sometimes contradictory) information to try and simulate the old plates. After digesting all the info, filtering out the misconceptions and making some educated decisions, I came up with a new orthochromatic template and one for early panchromatic plates. You could just go and buy pre-canned Photoshop filters, but where's the fun in that.

 

Here's a starter for 10 - a pic of two of my kit built wagons, put through the two monochrome templates:

 

I'll dig out some of the original pics I used for calibration to show my methodology. I'm not claiming the filters are 100% accurate, different companies plates/emulsions varied in the results you see and this is just my best attempt at a simulation. I've also got some other pics (not filed very well) that show just how confusing colour interpretation can be.

 

Post a pic and I'll run though the templates!

Edited August 5, 2023 by 57xx
Correction of terms

[magmouse]

Thanks, Ric, for starting the thread, coming out of discussions between us and a few other forum members on the thorny topic of GWR wagon colours, and more widely how colours appear in black and white photos of the 19th and early 20th century.

 

For me, I have an interest that extends the topic beyond the specifics of trying to rediscover wagon liveries, and to a more general question of why photographs of the above period look the way they do, and why it is so hard to recreate that aesthetic now. I am sure for those of us who model the pre-grouping period, part of the allure is the atmosphere, not just of the past as it was, but how it looks in the photos of the time. As a hobbyist photographer, and one-time professional theatre lighting designer, as well as a modeller, I am especially curious about visual aesthetics, what creates them in terms of techniques and technologies, and how we feel about what we see.

 

So, when Richard mentioned he had developed some specific methods in Photoshop to reproduce the look of old (ortho and panchromatic) film, I was interested to see how close that could get us to the look and feel of historic photos. As a starting point, we might experiment with this (public domain) photo:

 

Photograph by Hugh Llewelyn, licensed CC BY-SA 2.0. Original at: https://www.flickr.com/photos/camperdown/14420542493/in/album-72157627912565553/

 

 

Nick

 

[Compound2632]

Just now, magmouse said:

I am especially curious about visual aesthetics, what creates them in terms of techniques and technologies, and how we feel about what we see.

 

Physical differences will come into play, I think, including: aperture (small) and consequent depth of field; plate size and emulsion resolution (how does this compare with a modern digital camera); and (pincushion) distortion due to the use of simple lenses.

[57xx]

9 minutes ago, Compound2632 said:

 

Physical differences will come into play, I think, including: aperture (small) and consequent depth of field; plate size and emulsion resolution (how does this compare with a modern digital camera); and (pincushion) distortion due to the use of simple lenses.

 

There will definitely be aspects other than monochrome conversion we need to consider. I did do something on replicating film grain, I will have to look at that again. We can have a play with these other aspects too.

[57xx]

21 minutes ago, magmouse said:

Thanks, Ric, for starting the thread, coming out of discussions between us and a few other forum members on the thorny topic of GWR wagon colours, and more widely how colours appear in black and white photos of the 19th and early 20th century.

 

For me, I have an interest that extends the topic beyond the specifics of trying to rediscover wagon liveries, and to a more general question of why photographs of the above period look the way they do, and why it is so hard to recreate that aesthetic now. I am sure for those of us who model the pre-grouping period, part of the allure is the atmosphere, not just of the past as it was, but how it looks in the photos of the time. As a hobbyist photographer, and one-time professional theatre lighting designer, as well as a modeller, I am especially curious about visual aesthetics, what creates them in terms of techniques and technologies, and how we feel about what we see.

 

So, when Richard mentioned he had developed some specific methods in Photoshop to reproduce the look of old (ortho and panchromatic) film, I was interested to see how close that could get us to the look and feel of historic photos. As a starting point, we might experiment with this (public domain) photo:

 

Photograph by Hugh Llewelyn, licensed CC BY-SA 2.0. Original at: https://www.flickr.com/photos/camperdown/14420542493/in/album-72157627912565553/

 

 

Nick

 

 

Here is the Iron Duke in Ortho:

 

[57xx]

With added film grain and taken the edge off the focus.

[magmouse]

By way of comparison:

 

http://www.midlandrailway.org.uk/derby-registers/DY262

 

Some things to think about:

 

 - in the period photo, the sky is the only part of the picture 'blown out' (over-exposed). None of the other tomes get close to white - the closest are the areas where the light of the sky is reflecting in the shiny boiler of the loco behind the main subject.

 - there are some areas of complete shadow/darkness (possibly exaggerated by the digitisation of the original print or negative), but the overall effect is of a range of many mid-greys (50?).

 - The buildings in the background are lower in contrast - I see this in many photos of the period, and can only assume it is due to atmospheric pollution. I don't see how it can be a photographic effect, though it is an important part of the 'look'.

 

1 hour ago, Compound2632 said:

Physical differences will come into play, I think, including: aperture (small) and consequent depth of field; plate size and emulsion resolution (how does this compare with a modern digital camera); and (pincushion) distortion due to the use of simple lenses.

 

Yes, indeed - though with depth of field, the small aperture (greater depth) will be countered by the larger negative size (6.5 x 8.5 inches for 'full plate'). A large format in film will beat the highest resolution digital cameras made for general use. As for lenses, I don't notice much distortion in old photos - I think lens design got reasonably good quite early from that point of view. What matters more will be the micro-contrast (contrast at the very local level) - all the subtle stuff lens geeks get excited about. Old lenses certainly have a distinctive 'look'.

 

Nick.

[magmouse]

13 minutes ago, 57xx said:

With added film grain and taken the edge off the focus.

 

Now that's getting is somewhere. I think the sky should be white - with blue sensitive films, the sky is almost always white, with no cloud detail. The trees in the background are looking especially good. 

 

Nick.

[57xx]

Its the film grain overlay that's darkened the sky in combination with what is probably a touch too little blue sensitivity still. I will tweak more and report back! 

Edited August 4, 2023 by 57xx

[Mikkel]

Thanks for starting the thread and reposting your experiments, good idea to give the topic a thread of its own.

 

9 hours ago, 57xx said:

Post a pic and I'll run though the filters!

 

If and when you have the time, here are some model photos with wagons in the middle distance, which is how we often see them. I'd be interested to see how the greys and reds (and green, and dark and light stone) compare with your filters (although obviously me and my camera's limitations etc will affect the outcome).

 

 

 

 

Edited August 4, 2023 by Mikkel

[Compound2632]

8 hours ago, magmouse said:

As for lenses, I don't notice much distortion in old photos - I think lens design got reasonably good quite early from that point of view.

 

One notices it pretty quickly when one's focus of interest is the stuff round the edges of the image, not the ruddy great engine someone inconsiderately parked in the way.

[magmouse]

8 minutes ago, Compound2632 said:

One notices it pretty quickly when one's focus of interest is the stuff round the edges of the image, not the ruddy great engine someone inconsiderately parked in the way.

Do you have examples?

[Compound2632]

11 minutes ago, magmouse said:

Do you have examples?

 

The one that springs to mind is the third photo here:

 

 

See also the subsequent comments. I'll see if I can find the full photo.

 

But I've seen plenty of examples; often the wagon on the very edge of the plate looks longer than one further in, though it's the same type, which I think is what one expects towards the corners with pincushion distortion.

Edited August 4, 2023 by Compound2632

[Ian Smith]

First, May I also thank you for starting this thread and for the effort involved in your experiments.

Secondly, the thing that jumped out at me with regard to the Iron Duke photo experiment is the tone of the frames - in colour they are clearly polished brass (as are the tops of the splashers), however when the filter is applied the splasher tops still look polished bare metal however the frames look completely different looking painted rather than bare metal!

Ian

 

[Chas Levin]

Morning all, as others have, may I say thanks for starting the topic as it's something a lot of us spend time discussing and wondering about.

 

There's been some discussion recently about the question of how red, black and white appear on the different early types of film in threads on LNER Sentinel railcar livery and on the Sonic Models GCR/LNER A5 - happy to add links if anyone can't find them. Both threads had examples to illustrate, but perhaps lack of examples isn't the issue, it's correctly interpreting the many that we have!

[billbedford]

I can think of a couple of experiments I would like to see. One is the effect of titanium white v lead white and the other is how wagons painted (oxford) blue render on ortho film. 

[magmouse]

54 minutes ago, Compound2632 said:

The one that springs to mind is the third photo here:

 

Hopefully David won't mind me reposting the photo in question here, for easy reference:

 

 

A couple of things to say about this picture. Firstly, it is a photo of a print of a photo, clearly not lying flat on the table and taken at an angle, which makes it hard to judge what is going on in the original image. Secondly, it looks like it is a crop of a larger photo, which will tend to magnify any distortion or other artefacts that are present. Thirdly, we are looking at an angle at the row of wagons, so not only do the nearer ones appear larger, we also see them more flat-on. That will have the effect of the Hanwood wagon appearing longer than the MR wagon.

 

Overall, I don't find this a convincing case of lens distortion. For the period we are mostly looking at (say post 1885), low-distortion lenses were well established - see for example the Rapid-Recvtilinear of 1866 -  https://en.wikipedia.org/wiki/Rapid_Rectilinear. See also https://en.wikipedia.org/wiki/History_of_photographic_lens_design if you are interested in the wider context of lens development.

 

1 hour ago, Compound2632 said:

on the very edge of the plate looks longer than one further in, though it's the same type, which I think is what one expects towards the corners with pincushion distortion.

 

We need to distinguish between actual distortion of this type (radial distortions, which include pincushion and barrel distortion) and so called perspective distortion, which isn't really a distortion at all, but the natural effect of projecting a 3D world onto a flat image plane.

 

Nick.

 

[Compound2632]

1 hour ago, magmouse said:

A couple of things to say about this picture. Firstly, it is a photo of a print of a photo, clearly not lying flat on the table and taken at an angle, which makes it hard to judge what is going on in the original image. Secondly, it looks like it is a crop of a larger photo, which will tend to magnify any distortion or other artefacts that are present.

 

I acknowledge these points; indeed I had thought a scan had been posted but that may have been lost in the great crash. It is an edge of plate crop from a photo of a Web 3-cylinder compound, at Shrewsbury shed.

 

I can probably find better examples but after I'm back from holiday! 

 

Anyway, this peripheral, so to say, to the main subject at hand.

Edited August 4, 2023 by Compound2632

[magmouse]

4 minutes ago, Compound2632 said:

Anyway, this peripheral, so to say, to the main subject at hand.

 

Are you saying this rabbit hole has a rabbit hole???

[Jeremy Cumberland]

2 hours ago, Ian Smith said:

 

Secondly, the thing that jumped out at me with regard to the Iron Duke photo experiment is the tone of the frames - in colour they are clearly polished brass (as are the tops of the splashers), however when the filter is applied the splasher tops still look polished bare metal however the frames look completely different looking painted rather than bare metal!

Ian

 

Surely the frames are painted brown? I have never seen the locomotive, but the frames definitely look to me to be painted brown in the colour photograph, the same colour as the edge of the front buffer beam. The axleboxes might be bare dull brass, but I am more inclined to think they are painted too, in a lighter brown than the frames.

[Ian Smith]

1 minute ago, Jeremy Cumberland said:

Surely the frames are painted brown? I have never seen the locomotive, but the frames definitely look to me to be painted brown in the colour photograph, the same colour as the edge of the front buffer beam. The axleboxes might be bare dull brass, but I am more inclined to think they are painted too, in a lighter brown than the frames.

Apologies, you are quite right. I originally viewed the post on my phone (so the image was quite small). Looking again at a much larger image reveals that what I saw as a brass shade is indeed painted a brown colour.  Being a post 1900 modeller I am used to seeing (or imagining I’m seeing) Indian Red frames, Iron Duke is clearly sporting a lighter brown rather than Indian red.

Ian

[57xx]

2 hours ago, billbedford said:

I can think of a couple of experiments I would like to see. One is the effect of titanium white v lead white and the other is how wagons painted (oxford) blue render on ortho film. 

 

I think the blue should be fairly easy to show, although blue response is one of those things where there was a lot of conflicting info. What are you looking for in the titanium white vs lead white?

[57xx]

4 hours ago, Mikkel said:

Thanks for starting the thread and reposting your experiments, good idea to give the topic a thread of its own.

 

 

If and when you have the time, here are some model photos with wagons in the middle distance, which is how we often see them. I'd be interested to see how the greys and reds (and green, and dark and light stone) compare with your filters (although obviously me and my camera's limitations etc will affect the outcome).

 

 

 

Here's a quick rendition of your first pic. I think I need another tweak on the hue/saturation. Looking at my calibration pics, things seem to have wandered from where I had them originally! I've done a bit of softer focus and taken off the film grain layers, that seems to be giving some odd results, so I need a better way of doing it.

 

[57xx]

12 hours ago, magmouse said:

Some things to think about:

 

 - in the period photo, the sky is the only part of the picture 'blown out' (over-exposed). None of the other tomes get close to white - the closest are the areas where the light of the sky is reflecting in the shiny boiler of the loco behind the main subject.

 - there are some areas of complete shadow/darkness (possibly exaggerated by the digitisation of the original print or negative), but the overall effect is of a range of many mid-greys (50?).

 - The buildings in the background are lower in contrast - I see this in many photos of the period, and can only assume it is due to atmospheric pollution. I don't see how it can be a photographic effect, though it is an important part of the 'look'.

 

This is proving to be quite a challenge (not a bad thing!). By adding film grain, I lose the blown out appearance of the sky. It appears to be a bit tricky getting that over exposed sky just right, even bleaching out the blues so no clouds appear. I can still just see the clouds on the Iron Duke pic. You sample period picture really captures the elements we are after well. I'm wondering if the buildings haziness is also in part down to the over exposed sky bleeding across? Not sure if that is even "a thing" that happens on the old film. Certainly the more distant objects have that colour fade with distance effect from the atmosphere.

 

There is a lot going on in the reference pic that will take some thought to replicate. 

[Mikkel]

2 hours ago, 57xx said:

 

Here's a quick rendition of your first pic. I think I need another tweak on the hue/saturation. Looking at my calibration pics, things seem to have wandered from where I had them originally! I've done a bit of softer focus and taken off the film grain layers, that seems to be giving some odd results, so I need a better way of doing it.

 

 

Many thanks. Hard to tell the dark grey MSWJR from the GWR red wagons. The D299, on the other hand!