Scale Modelling

[chris46]

Scale models, regardless of the category they belong to, should be a lifelike, scaled-down replica of an original (prototype). This means that I collect reference materials (photos, descriptions, color schemes, etc.) that accurately depict the subject of the future model.
Unfortunately, I repeatedly find that some modelers do not conduct their research very thoroughly and sometimes use documentation from other prototypes that do not correspond to the same subject. Whether the model was built "to scale" down to the smallest detail is not important in this context. Ultimately, everyone can decide for themselves how far they want to go with the detailing. What is important, however, is that the accompanying materials—photos, color schemes, descriptions, etc.—correspond to a single prototype, especially if a report on it is to be published.

Chris

[g_kandylakis]

This is a universal truth, when it comes to scale modeling, especially if we are talking about competition.

The normal course is as you describe, search - find - document - build...

Many people finish a model first, fly it, like it and want to enter it in competitions. So they then search for the necessary documentation, which of course almost never corresponds to the particular model, especialy if it was built from a kit or old plan.

Which brings to mind one of my favorite pink panther cartoons, about the proper order...

George

[Stunthenk]

Again and again, in spite of the Pink Panther, the main problem we find when researching a subject from the early decades of aviation, is NO COLOUR DESCRIPTION! Pilots and technicians from the earliest era to modern times are not interested in the colours of their machines. Only the specifications and abilities....

[Jmk89]

Again and again, in spite of the Pink Panther, the main problem we find when researching a subject from the early decades of aviation, is NO COLOUR DESCRIPTION! Pilots and technicians from the earliest era to modern times are not interested in the colours of their machines. Only the specifications and abilities....

True for non-military types, although before WWI natural linen seems to have been most common.  For WWI OG Thetford's book on WWI camouflage is a good start - available in the Plans Gallery.

Otherwise, this is an example where, if usual sources aren't very helpful, it is useful to use ChatGPT or equivalent.

For example, if you ask ChatGPT about the colour scheme of the Sopwith Atlantic (where I have previously tried to find out information about colour scheme), you get this:

"If you mean the 1919 Sopwith Atlantic, the surviving photographs and contemporary material suggest a fairly simple finish rather than a colourful squadron scheme. It was a one-off experimental aircraft built for the transatlantic attempt.

For colourising your photograph, I would use approximately:

Upper surfaces: medium/dark olive-green/khaki protective dope
Fuselage: similar khaki/olive-drab, with some areas appearing lighter due to the fabric finish and lighting
Lower surfaces: very light cream/off-white or pale doped fabric
Engine cowling: dark metallic/blackened metal
Wooden propeller: natural medium brown, with darker laminated sections
Struts and wooden framework: natural wood or light brown
Metal fittings: dull aluminium/steel
National markings: British-style roundel markings where visible, but don't introduce bright modern RAF colours unless supported by the photograph.

There is also contemporary documentation specifically referring to a "Sopwith Atlantic Scheme" with four views of the aircraft, which is useful evidence for reconstructing the finish."

That actually makes sense - the original was a variation of a B1 so it was probably taken out of the spare production models that were around, which would have been finished in RAF colours, and Sopwiths probably had a few gallons of khaki dope going spare at the time!  My only query would be whether the rear top fuselage would have been doped or fabric covered at all.  My recollection is that Harry Hawker made a boat which fitted behind the cockpit instead of the usual stringer arrangement.  In that case if might have been plain varnished wood - like the prop.

You can then ask Chat GPT to colourise some B&W photos.  When I did this I get the attached suggestions, using photos from Google images (BTW the labelling on one photo is wrong about the cockpit, but none of it is mine).  The boat area in the labelled photo does look to be stringered and fabric skinned rather than all wood - which also makes sense, since weight was an issue (hence the drop-off undercarriage).

[billdennis747]

It's always been fun reading interpretations of B and W photos for colour, and it looks like AI will be adding a new level to the amusement (the engine cowlings were blackened metal?). There are a lot of photos scattered about here about the boat, and the top decking is clearly plywood (it's the boat!) but there is a confusing reference to fabric. I'm not even sure which is the pointy end.

https://www.kingstonaviation.org/sopwith-day-by-day/1919-2/

[Stunthenk]

Unfortunately, the BMFA scale rules on proof of colour (6.1.2.3.b) say: Correct colour may be established from colour photographs, from published descriptions, from samples of original paint or from published colour drawings. For classes CL1, RC1, FF2 and FF3, writen descriptions must be accompanied by colour chips certified by a competent authority'. And under Penalties for inadquate documentation (6.1.2.4 d): Inadequate proof of colour: ZERO points for colour!
The only escape clause for pre-colour era photographs and lack of contemporary descriptions is 'published colour drawings'. Which is a bit strange, considering the many cases of incorect colour drawings published over the years!

[billdennis747]

That's why the rules give so few marks for colour, they can almost be ignored!

[g_kandylakis]

Quite right Bill. 2% of the total static score is color accuracy. Almost nothing in total.

But for me there is another complication when it comes to missing color information.
I simply cannot get myself to build it knowing I do not know the color...

I do feel we have drifted a bit off topic though. Chris's original post was about bad or incorrect documentation. Missing color information is only part of that and beyond the control of the modeller in many cases. So not as bad..

[g_kandylakis]

As for AI, my initial reaction is, forget it...

It is not AI at all, just a thorough collection of everything that is online, filtered and presented in a nice way. If the initial info is wrong you get nonsense.

Same with picture coloring.
I uploaded a picture of the Elias Aurcoupe, a documented YHWM, only to get back a dark green wing with US insignia...