It’s a model. Deal with it.

When I look at a painting, the last thing I want to see is an exact re-location of real life: I want something of the artist to show through, and to be able to try to understand what the artist is showing me. I am happy to accept that this is a painting. I don’t expect it not to be. When I look at a good photograph, I know that what I am shown is dependent on the viewpoint (literally) of the photographer and the lenses used, and that is part of the appreciation. I don’t attend live music expecting to hear what was recorded in the studio, and I know that the play is an artifice.

When I look at an individual model, I personally look for a fairly authentic replication of the real thing, to within a certain (hopefully specified) tolerance of real dimensions, with believable colour rendition and clean workmanship. When I look at a layout, however, precise replication of the real place can be problematic, and sometimes frankly boring, so I look more for the interpretation of the scene, the composition of the model, and the consistency of the craft used: items like track clearances stand out to me, but I can cope with (say) 1:8 turnouts in place of 1:10, as they often look sharper in real life due to seeing track from just a few feet above ground level.

My models use electricity to run them, not steam. I have to manually couple and couple them, and I don’t tighten the shackles on screw-links. I could use auto couplers, but the real thing didn’t – even knuckle couplers need lining up manually – except on some multiple unit stock, which I don’t model.I can’t do anything about the periodicity of loose couplings swinging, as that is down to their length and has nothing to do with mass, and so on.

In short, it’s a model. I know that. I don’t mind that. I just want it to be vaguely recognisable as something prototypical and which reflects the effort that went into its creation to make it as un “toy-like” as possible.

And yes, when I see train set track nailed to a piece of plywood, which has been painted grey to represent ballast, playing host to unmodified ready to run trains, rushing past plastic train set buildings and through dyed sawdust fields, I generally take the view that not much effort went into creating it, and pass by before I start to feel insulted.

Do you get it?

Chances are, if you are in the very small minority of modellers who follow this blog, you probably do get it. Maybe I am whistling (or worse!) in the wind, or preaching to the already converted (or choir, if you are North American).

Anyway, I recently had the temerity to suggest that all the effort of making a layout with correctly gauged and highly detailed track was for nothing if a loco running on it wasn’t well-enough proportioned/shaped to capture the look of the prototype, even if it was nicely weathered. As the layout concerned has been built to “Protofour” standards by a member of the Scalefour Society, I even quoted their frequently used strap line: “it’s not just about wheels and track”.

The responses are, if you are not involved, an amusing display of self-righteous priggishness, not too far removed from the indignation a letter in the Railway Modeller created, when it was suggested that a minor detail on a layout was out of period, and the required information for correcting it was helpfully provided. I have been told I am wrong, I am petty, that the model concerned looks like the prototype, it runs really nicely, and that the weathering job makes up for it. And oh yes, that “P4 is easier once you realise not everything has to be precisely accurate” because proportion and colour are more important. Let’s go through those points.

Firstly, it was a genuine question, not a direct criticism: what is the point of going to the nth degree of accuracy in one part of a layout, if the rest isn’t at least proportionally correct enough to create a convincing display of verisimilitude? Secondly, if someone has gone to the trouble of making their own track and replacing the wheel sets on their models in an attempt to (presumably) make better models, is it not odd to then run something which is a bit of a caricature of the real thing on it? Thirdly, the model is inaccurate enough in two areas (the under frame of the prototype does not have pseudo-solebars running under the length of the body side, and the cab roof curvature is too shallow) to have prompted several articles on improving it and a new manufacturer to go into production solely to rectify the errors in the models of the whole series of related prototypes. Fourthly and fifthly, a nice weathering job and good running are irrelevant to the matter at hand: these are prime examples of the “straw-man” fallacy. Finally, I never mentioned the need for the loco to be a precise, accurately measured model: just to look like the real thing. (Colour is another straw man.) The “precision” of the modelling is relevant only because the layout builder has chosen to follow “finer standards” in his modelling, and I wondered about the consistency of the approach.

At the same time, in another part of the same bulletin board, a comment was made about S scale modellers being concerned over as little as 1/64th of an inch. This amused me greatly, for two reasons. Firstly, in the UK we use a scale wheel profile and track gauges deal with laying track accurately: in this respect, the precision of the modelling is as good as the tools (and hence usually, a lot finer than 1/64th of an inch!) Secondly, getting things to within a scale inch or so is simply the easiest way to create a model with the correct proportions: getting the proportions and relationships between objects is more important than precise “exact scale” accuracy. Working to a high level of engineering tolerance is a means to an end, which is something that looks good and runs well.

Those who have read the S Scale MRS history page will know about Charles Wynne defining his level of accuracy for modelling: within a scale inch of the prototype dimensions. To me, that is what “finescale” is all about: defining your tolerances, and sticking to it. He did this 1919!

Which brings me back to my main point, and also takes me back – yet again – to the words of Cyril Freezer in the mid 1970s, which gave guided my thinking on “toy” vs “scale” vs “finescale” ever since, but acknowledging improvements in manufacturing processes since then, I am going to paraphrase things a bit – and besides, these are my thoughts on the matter: fundamentally the same, but refined in the light of 46 years of thinking, modelling and improvements in the quality of toys.

If you assemble a model railway using off-the shelf components, and base the layout around off the shelf “universal” track, then you are essentially trying to create either a toy train set using modern, well-made and reasonably accurate models, or are trying to create a scale model railway where the overall impression is of paramount importance. There are many reasons for doing this, and if this is your choice, then it is correct for you even if it isn’t how I would do it. I will be the first to support your choices and the last to apply my standards to your models. In this regard, the word scale is being used to say that the models look more or less like the real thing (“near enough is good enough”), and are reasonable enough representations of the prototype to be a model in that sense: i.e. this is a model railway, not a toy. Good luck to you: enjoy it. And I will be happy to operate your layout if asked.

However, once you decide to start move away from the out of the box experience, possibly with a bit of detailing (replacing cast/moulded handrails with wire, for example) then you are stepping away from this comfort-zone. This is particularly so when you start to address the issues of running quality and compromises required to produce models to “universal” model standards. It doesn’t matter if you are working in 4mm scale, and simply refining 00 track, replacing wheels snd moving up to EM or P4 track (even if it is the new Peco EM track) or work in H0 scale, and are using “code 88” or “code 55” wheels, and replacing trucks with better looking, more correctly proportioned upgrades, or dealing with similar issues in other scales (including North American S scale!). If you have decided to create models that are closer to the prototype in terms of colour, shape, proportion, fine detail, dimensional accuracy, or track which looks like the real thing reduced in size, then you are modelling to “finescale”, and getting things consistently right is important. If you think that an accurate track gauge is required, then displaying models with an incorrect shape is pointless, no matter how good the “weathering” or the running qualities. Once you know, it has to be done…

If you can’t see this, then you really don’t get it at all.

DCC: getting long in the Bluetooth?

I thought that, ten and a half years after originally penning this, I would repost it.

Yes, there are now off-the-shelf solutions from the excellent Tam Valley Depot for both BT and wireless communications, but compared to the small thumbnail devices I see around, they are bulky. Granted, there is a need to combine this with battery charging/regulation and current step up, but there are small circuits available for doing this.


Thinking aloud, and knowing it is technically possible, but not sure how, how about an idea for the next stage in DCC?

Conventional control systems apply 0-12v DC over the rails, and any locos not isolated respond to this and they move. (Also multiple units, carriages and what have you, but I would like to keep this simple.) The rails supply power and control, as a single combined form of energy.
DCC has a constant AC voltage for power, and superimposes on this an encoded command signal which is received by all locos, but is ignored except by those to whom the command is directed. This is all done over two-rails, which serve the purpose of providing power and also of conveying the command.

Most wireless DCC systems remove the tether between the base station and the individual control units, but the loco still receives both power and control over the rails.

Now, there are some interesting alternatives, such as Locolinc and also CVP’s Airwire900. Locolinc is a proprietary system, and that’s about that, really. Airwire900 is great, but like Locolinc is not suitable for use in the EU/UK due to the frequencies it uses. They both have the right idea: battery supply of power and wireless transmission of the control signal. But if everyone turned up with one of these systems at a model railway exhibition, we might get haywire rather than airwire as I am unclear on how they pait the control units to the locos. Much closer to an ideal is the do-it-yourself approach of the Aussies, with the concept of DWiDCC (Direct Wireless DCC) but again, this is using radio frequencies. Both DWiDCC and Locolinc point out that one could use the track to provide a trickle charge for the on-board batteries – when it came to reverse loops one would simply have a dead section longer than the longest loco, and point crossing vees simply do not need wiring up at all. (And the power provided could double-up for track circuiting purposes.) These systems are all, in their own ways, brilliant and yet…

…and yet, why, instead of this direct-to-loco control via radio frequencies, can we not have a simple bluetooth setup?

What I have in mind, is an interface which plugs into my DCC command unit. This makes the base station a key unit with overall control (there could still be a programming tack or output if so required). Throttles can be tethered to this, or themselves could plug into a bluetooth transmitter unit. And for locos, a simple bluetooth receiver, into which power is fed from the batteries, and which superimposes the DCC command signal over this before feeding into any NMRA-compliant DCC decoder you care to think of.

If provided as a series of bluetooth components, for the base, throttles and locos (and accessory decoders – why not?) the system is independent of all others, and anyone can then turn their tied down system into something really revolutionary. With modern motors (not just coreless, Sagami and Mashima, for example, are efficient) and advances in battery technology, power is not an issue.

So, if anyone out there interested in DCC knows enough about blue tooth to make this work, please, please design a prototype and let us know about it!


Also, why not call it BCC, Bluetooth Command Control?

“Once you know, it has to be done”

The above is a wonderfully resonant quote from a model railway forum post.

It refers to the fact that the builder had made an assumption about a feature on his chosen prototype, but when he got around to checking up at his local club’s library, he found he needed to make some corrections. (This is a very minor example of the disruption caused by COVID-19, as he didn’t want to put the project on hold pending all the information.) Not everyone would have bothered, but that, perhaps, is the most succinct and compelling difference between a genuine “finescale” model-maker and someone who doesn’t want to get it right. I might add that wanting to find out these details is also a key part of the finescale approach.

There is also the point that pending full knowledge of prototype practice, a reasonable interpolation of the design was made – no hanging around waiting to “know everything”, just an acceptance that a correction would be made if necessary later. This applies to most things within our hobby, although once the track is laid, it can be difficult to change certain fundamentals of its design and construction without wholesale destruction!