Scenery design, compression, and illusion

From LdsigWiki

Jump to: navigation, search


Why Plan Scenery

The roar of diesel prime movers echoes from the canyon walls as a set of shiny General Electric thoroughbreds strains to lift their load over the verdant mountains. Far away, in time and space, a mikado stretches its white plume along its tail as it speeds manifest freight across the plains. A beleaguered switcher trundles into an industrial park shouded by the mist as it makes its morning rounds.

Most of us have some image of railroading that we want to capture with our layouts. The image won't be captured by passing sidings, yard ladders, car mixes, operating sceanrios, or even aisle widths. Indeed, while these elements are essential for the operation of the layout, they do little to capture the atmosphere that we seek. There needs to be something beside the tracks to give the viewer a clue to the location and type of railroading that is modelled. This missing piece of the layout is scenery and its design is the subject of this chapter.

Planning the scenery is important because while we all want the railroad to look like a reasonable representation of the prototype, we also want it to be fun to operate. Unfortunately, these two goals are often at logger-heads with one another.

There are at least two reasons for this conflict. First, scenery takes room. For example, a reasonable sized house is about the same width as two tracks; a large oak or maple treee is about as wide as three tracks, rivers take even more room and real hills and mountains are so big that they can't reasonably be modelled even in small scales. Thus, if we do not account for the scenery in the initial track-planning phases of design, we may find ourselves with no room for the atmosphere when it comes time to actually sling the plaster or plant trees. The result is a granger railroad full of retaining walls and cliffs, or a world where all the buildings are full of "interisting" angled walls.

The second reason why scenery and operations sometimes don't mix is that the world is just too big to model. Some of the illusions that we can employ to address this problem require constrained viewing angles. Either the scene must be close to eye level, which makes switching inconvenient, or view blocks need to be used which operators may need to reach around, or which occlude the trains during operation. Failure to consider these viewing angles at the onset may lead to the easy discovery of the illusion. The model world will not look as big as intended, or worse, it will look completely unrealistic.

Of course, the perennial tension between operation and scenery is not the only reason why we consider the scene early on in the design process. Another reason is expense. Scenery can be quite expensive both in terms of money and time.

Forests are perhaps the canonical example of this phenomenon. Particularly with the smaller trees that are common on our layouts, it takes an incredible number of trees to make a forest. For a pine forest typical of the Pacific Northwest, for example, there will easily be 100 trees crowded into a square foot of real estate. Even the fastest and cheapest imaginable methods result in a forest that is in the order of 25 dollars and three hours per square foot. Many scenery designers do not realize this simple fact, and consequently point to wide open expanses of benchwork in the middle of a loop and say, "this will all be forest." Little do they realize that a 30" radius loop will require 2000 trees to fill the middle.

Naturally, even more costly lessons greet us when we plan vast expanses of city-scape. So whenever time and money are serious constraints, we should try to minimize the costs of our decisions. After all, we are railroad modelers so shouldn't the bulk of our resources go to locomotives and rolling stock rather than foliage?

If practical considerations such as cost or operation are not enough to convince that it is a good idea to plan for scenery in advance, then consider the aesthetic reasons. Sculptors sketch their works before committing chisel to marble so that they can ensure that the final product will be pleasing to look at. So too should the model railroader, or risk having a layout that is a hodge-podge of scenes, lacking cohesiveness, and jarring to the viewer.

Besides, scenery design is fun.

The Process

The Real World

The first step in designing the scenery for a layout is the same as the first step in designing every other aspect of the layout: observe the prototype, Here we have an advantage over trying to replicate the operations or even the track plan of the prototype. Scenery hasn't changed all that much in the past hundred or so years. Some observing of the prototype of today will give many clues to what the scenery looked like in an earlier era.

While observing, we take not of every aspect of the scene. Where do trees grow? What types and how far apart are they? Are the rivers rushing torrents or lazy meanderers? Are buildings close together, and what shapes are they? How steep are the hills, and how big are the cliffs? All these factors and many more come into play when planning the scenery.

One aspect of prototype scenery that should be noted is that most of it is boring. After all, railroads tend to seek out the least expensive route, and it's expensive to cut a shelf into a solid rock wall. Even in mountainous territory, the line will tend to seek flatter areas. There is certainly spectacular scenery, but it is sprinkled sparingly throughout a great deal of drudgery.

Buildings obey a similar rule: they are square and made of unspectacular materials because it's cheaper to build them that way. Even when the line does pass close to a landmark, it often does not pass within the few hundred feet which is the typical depth of our benchwork. Our scenery will appear more realistic if there is not a "Phantom Rock" on every curve. We should learn to delight in the ordinary.

Once we have observed the world for a while, we will have a list of typical scenic elements that will look right on our layout. These elements might not come from the list in the second section, but could be representative of the region being model led. They might be a typical scene such as willow trees growing along stream bank, or they could be a typical building type such as the Ontario cottage, or even a landmark like the Tehachapi Loop. The task is then to arrange these elements in such a way that the effect is pleasing and realistic

Prototype Modeling

Before we go on to the next section, we should address the possibilities of prototype modeling. We might expect the scenery plan to be complete if a prototype scene is the starting point. After all, we have a plan and perhaps photographs from a variety of angles that show every bush, building, bridge, and heap of tie plates that existed on the real thing.

The problem with prototype scenery modeling is that it is impossible to model everything. Even if we can constrain our appetite to a very small swath of the world, we can't allow every single view of that world. The prototype modeler must choose which viewpoints are to be modeled and which are left out.

Iain Rice likens this process to photography. While a painter, like a freelance modeler, decides what to paint on their canvas, a photographer must choose the most interesting views of the real world. Of course many prototype modelers are unable to constrain their modelling to a small piece of the world, and in that case, the scenery design process is somewhere in between the carte blanche of the freelancer, and the rigid world of the purist. I guess people that are willing to bend the world in this way are something like a photographer with a dark room.


There is nothing ingenious about the scenery design process, or if there is, no one has shared it with me.

The central point is this: start early.

The track planning process typically involves sitting down with a big bag of chocolate chip cookies, some milk or other lubricant, and a great deal of paper and pencils and above all erasers. If during this sketching and eating phase we consider the scenery and the list of scenic elements that we came up with while observing the world, we will be more likely to find room for the scenery. we will include the scenery on the sketches.

Everything should be included on these sketches. Starting from hills, cuts and fills, I add buildings, then trees and bushes and even ground treatments. As we sketch, we will build a mental image of the layout as it will eventually appear.

Artistically inclined layout designers might take this opportunity to sketch scenes from the plan. They imagine themselves looking through the lens of a camera shooting their layout's feature article for some magazine, and sketch what they see.

Once we have a plan that we believe will suit our purposes, we should check some of the angles a little more carefully to ensure that the scenes we have in mind are going to occur. We take a ruler and draw lines across the layout from several possible viewing locations. In particular, draw lines from the end-most viewing positions to the opposite corners. These "sight lines" will show which features will be visible from which viewing locations.

We use the sight lines to ensure that the view blocks will work as anticipated. That is, they will hide the features that need to be hidden and expose the features that should be visible. For example, sight lines on the plan will quickly determine if the viewer can see the joint between the backdrop and a building. Similarly, they can be used to ensure that the fouling points of switches are in sight.

These sight lines are of course a simplification. Foreground objects do not occlude everything behind them unless they are at eye level. However, for most purposes, a few straight lines will expose a multitude of sins, and allow us to try various absolutions before committing to plywood or even plasticene.

Mock-ups and Models

Finally, if we don't trust our drawings and our mind's eye, we might make a small-scale model of the layout. Many people consider this "meta-model" to be scenery design. By this point, however, we should already have a very good idea of what the sceney will look like. If not, then the scenery model is really a reactionary test to see if the track plan can be scenicked.

While that is better than not considering the scenery plan at all until we have a pot of rapidly-curing hydrocal in hand, the improvement is marginal. The reasons for this somewhat heretical statement are twofold.

First, building meta-models or mockups takes time. A reasonably accurate of an 8x8 layout can easily take an evening to produce. With this sort of investment, we will be unlikely to tear it apart and start over the way we might if it were a few chicken scratches on a piece of paper.

The second reason why scenery models are of limited utility is that most are too small. Apart from hills and streams, most elements are too small to position accurately and to get a feel for how they interrelate. Building a larger model, or even a full-scale mockup is of course the answer. But the larger the model, the more time and energy will go into it

While scenery models and mockups are not a very good design tool, they are the quintessential communication tool. So, if the scenery design needs to be shared with the rest of the club, or the Layout Design SIG, a model is probably the best way to achieve understanding.

That being said, we should probably explore the construction of such a model. It starts invariably with a track plan mounted to cardboard. This is cut out to approximate the shape of the roadbed and any other flat areas. The roadbed is then glued to blocks that raise it the appropriate height above the floor or other base elevation. Finally, the area between tracks is filled in with appropriate scenery contours using plasticene or modeling clay, and structures made from wooden blocks are added.

Incidentally, the amount of green plasticene used in this process is a pretty good indication of the number of trees that will be needed. If we find ourselves running back to Toys R Us for more green, then we can be fairly sure that we are going to be making a significant investment in Woodland Scenics sometime in the future.

Scenic Elements


Animals belong in the scenery plan in so far as their presence makes a small area of interest. Note that when we travel by train, it is actually relatively rare to see deer bounding into the forest (well that's what I've found, but then I've never won a lottery either). So, one package of Prieser deer is about all you can probably use on a layout.


The backdrop is a graphical representation of the world beyond the layout. It is treated fully in the Section on Illusion.

Bridges - Rail

Bridges form a natural scenic centerpiece for a scene. A train bursting over a chasm, or trundling across a long low pile trestle is full of drama. We should note, however, that bridges follow the trend to modesty that the rest of scenic elements follow: there are far fewer large spectacular bridges than there are puddle jumpers.

Bridges - Road

Road bridges are an interesting, less frequently modeled scenic element. The use of a highway overpass as a scenic break at the end of a scene is, however, a little cliche.

















Forced Perspective


Gaining Apparent Distances

Hiding Exits




Hiding Supports

About this content:
Original author: René Gourley. Last revised on 1996.
This LDSIG article is ©1996 by RenĂ© Gourley (email).
Questions/comments may be posted in the discussion tab.

Personal tools