Friday, March 15, 2019

Cheap Trees Are Made of This

The following blog post isn’t about a specific recipe or a short cut to make el cheapo trees but rather a proof of concept. The idea was to create a quick Eastern pine tree and see how it could fit in various scenes and environment. In this regard, my goal was to replicate the shape of the tree and not the correct texture or color.

Not bad when seen from afar...
The tree itself was built in about 10 minutes using a wood skewer (not bamboo), small twigs and pre-flocked netting found at a local gardening supplier. I used the true and tried method of drilling holes into the skewer and inserting branches in it. As previously stated, I cared much more about getting the shape right than the colors.

Making this tree proved me a few things I didn’t expect. First, modelling conifers such as pine is indeed as fun as building a structure or a freight car, with a certain dose of artistic freedom. Second, when you get the basics right, using finer texturing methods, improving branches and needles and coloration aren’t that hard.

...but it doesn't hold up under closer scrutiny.

While making this tree was quite fast and yielded surprisingly decent results, it is evident ones would be better off making them with more care. The pre-flocked netting saves a lot of time (applying needles and painting) but the results is not that much realistic and if you start airbrushing more realistic colors, it defeats the gain in speed you are going after. It must also be noted pre-flocked netting isn’t the most geometrically realistic material to use to replicate branches.

Another word of caution is the tree looks quite good on pictures, but is not that great upon closer inspection.  It means depending on your available time and the context of your work, such tree could indeed be a solution. But in most of the cases, taking a few hours working on maybe a half-dozen well done trees would be much meaning in the long term.

Tapered trunk, bark texture, more realistic branches and colors are a few areas that would significantly benefit from improvement. Also, looking at real Eastern pine growing in a nearby park, I found out most of them didn't have a perfect trunk but were a quite asymetrical.

However, given I’ll have to create a very dense forest on the layout, small trees made of netting that are to be hidden and embedded within the vegetation without protruding over the canopy could be a solution to bring some colors among leafless deciduous trees.

OK for background use? Not so sure...

Anyway, point in case, modelling trees is much more interesting than I thought and relaxing too. Given I’ll probably have to create about a dozen of pine maximum; I’m ready to do a better job than trumpet about a half-baked solution.

Thursday, March 14, 2019

Clermont: Terraforming... again

It seems to me not so long ago I wrote about terraforming Clermont... both here we are again, doing it for a second time.

The tree ridge is about the hill's planned final height

This time, it is going quite faster since we know what we are doing and all club members are actively participating in a manner or another in the process.

Some subtle landforms sloping down toward the valley

As you can surmise, last Saturday, we started the day by studying photographs taken last year and from there, we were able to draw on the plywood the various landform to replicate. We were quite bold in determining the size of the small valley and I'm curious to see how it will turn out. However, I believe the scene unusual depth in this part of the layout won't be bad at all.

Hiding cars and tracks on the other side of the peninsula is challenging!

While I'd like to make you believe this process is going straitghforward, I prefer to tell the truth behind this process. After a long day of work, we took a step back and tried to assess our progress. While some parts are genuinely close to the prototype, it became quickly clear our hillside profile was somewhat wrong. After a few inches of height, the slope should start to become gentler (about 35 degrees) until it reaches the summit. Unfortunately for us, we made the hill ridge a little bit shallow which means we will have to enlarge it of a few inches more if we want our summit to stay at the same height. No big deal, but we feel it must be addressed before moving forward with scenery.

The summit height has also a big implication on how we perceive the scenes together. The previous mountain was high enough to hide the tracks on the other side of the peninsula. However, this is no longer the case with a 6" summit (about 40ft in real life) and we will have to rely on vegetation to get another 6 inches in height.

A proof of concept for a decent "opaque" forest

To make sure it would work, we created a mockup forest reusing old twigs and supertrees from the previous scene. However, this time we tried - as best as we could - to get the tree density right. On most layout, this is overlooked quite a bit because trees have leaves. In our case, we need an opaque screen of leafless branches.

To get the effect right, we played on several parameters. First, the tree density which is a 3/8" (3 feet) spacing between trunks. It may seems overkill, but dead twigs are free and plenty. When painted the right color, they can help to fill a lot of ground.

Second, a few supertrees are mixed within the forest to create a denser core of branches. Supertrees' fine branches work wonders in that regard, including bushes and smaller growth. When planting trees, we will have to create several layers of various types of trees and it will be important to make sure they are planted in a realistic manner.

Third, depth. By depth litterally mean scene depth. Our mockup is about 4 inches deep and is already almost opaque. This means tree density will vary depending on location. Basically, vegetation on the ridged (where the forest depth is shallow) will be quite dense, but in the valley, it will be significantly reduced.

Fourth is color. Some twigs were already painted light gray while others were plain natural brown. Brown looks unrealistic for tree bark while gray looks far better, idem for supertrees fine branches. Another interesting point is that light gray is a light color. It makes elements look bigger and denser because it scatters more light. Thus, this is a perfect color to create an optical illusion of density.

I'm well aware covering about 20 square feets with so many twigs is ridiculous, but as previously mentionned, we have many parameters to finetune the results without losing sanity. Also, it is quite a relief to think these trees requires on very limited preparation (some trimming, Modge Podge treatment, painting and planting).

However, as Jérôme rightfully remarked, a few very well done trees will still be required to make the forest pop up a little bit. It includes a few conifers (I'll show you in a future post how I made a quick and cheap Eastern pine tree mockup) and a few larger deciduous trees near the track. The goal is now to create a multicolored patchwork, but rather a subtle variation in colors and textures with some elements bringing life to the composition.

Thursday, March 7, 2019

Clermont: Where Do We Go Now

I've been requested to clarify the new concept for Clermont. I can understand a cleared slab of plywood gives little indication about a master plan so here are a few ideas I wish to develop over the course of this year.

Analyzing the prototype from an actual survey

The decision wasn't made overnight and the new layout plan has been on the work since last fall. Interestingly enough, I found out old drawings I shared back then with Jérôme. They are a proof of concept and were meant to figure out if it was possible to create a better representation of Clermont.

Scaled down satellite image is a perfect fit

To make sure my intuition was backed by reality, I simply used a satellite image and reversed it (our layout is a mirror image of the prototype). Fortunately, it fitted perfectly our space and proved we won't have to fudge elements. That last thing was important to me because the moment you start playing with distance between structures, roads and other elements, realism goes down the drain.

A close up of Clermont

 As you can see, on Clermont side, all roads will be redone to closely replicate the prototype, including the intricate sloping streets and stone retaining walls. Topography will be much more gentle than it used to be according to pictures we took during our few visits on site.

Updated track plan

On the other side of the peninsula, a deeper scene will be developed around a small meandering brook. I already presented this idea last week, but I feel a plan better show how this scene will probably be about 24" deep. While convenient from a scenic perspective, this is also a way to save on florist foam and get the best out of available space. A large mountain top as we used to have is a pure waste of space. You only see the first few inches, the rest is out of sight. For this reason, it is a better idea to set a large valley where the peninsula is larger than a huge hill.

And let's be honest, building a shallow hill instead of a large mountain will take far less time and resources. That may sound lazy, but I mentioned it countless times our project has a lot of constraints we can hardly control. If sculpting a hill takes one week instead of a month, it means we can achieve much more and see a decent amount of progress. On the positive side, it also means we have more time for actual scenery (colors, textures) which have a greater impact on a scene.

Monday, March 4, 2019

Gone with the Wind...

Last Saturday, we cleared all the remaining obstacles impeding Clermont yard reconstruction, i.e. the mountain. I didn't shed a tear and neither my fellow club members. Meanwhile, Julien and André-Pierre paid us a visit and probably didn't expect to see the layout in such a state. I'm pretty sure they thought to themselves we were absolutely crazy! We will see... they may be right, maybe not!

Ready for the dumpster

That huge mountain was a neat scenery feat back in the early 2010s when we built it, but it looked out of place. Several times, we removed layers and cliffs to make it looks more natural to no avail. At the end of the day, we reach that point when keeping it was the proverbial incarnation of diminishing returns. One hour later, the entire peninsula scenery work was ripped off. Only the largest chunk of costly florist foams were salvaged for future use.

A clean slate!

From this point, we quickly sketched the new topographic features on the plywood. We will have to reinterpret Clermont landforms to better fit our space. While our yard is exactly the same lenght as the prototype, it doesn't mean locating the small brook in the exact spot will look right. However, I'm pretty confident it should look quite good and the scene will have more visual depth than before.

Shaping the right of way

The yard roadbed shoulders have also started taking shape. They were made from pre-shaped foam blocks following a 1:2 slope as seen on old CNR roadbed section drawings. It makes for a much more natural look than the previous 1:1 slope we used everywhere else.

They thought they would operate!

Finally, a nice diner with friends followed a productive day. We were all left wondering what expect from Canadian National's centennial celebrations. It certainly don't look like it will be a blast at this point, but the year is only starting.

Sunday, February 24, 2019

Wieland: More Terraforming

Another productive day in the bag and encouraging results to share.

A mockup of what we are trying to accomplish, note the subtle terrain shape in the background.

First of all, the terrain is now almost completely shaped from the furnace room up to Desbiens Street. We haven't yet worked on the yard topography because this is quite a large endeavour in itself and we feel it is better to complete work in Wieland.

Subtle terrain variations help anchoring the roadbed into topography

After scrutinizing a few prototype photo, I noticed some topographic elements I missed last time. It included more raised terrain near Desbiens Street that makes the track visually sinks into the scenery. Also, some small bumps of earth could be noticed near the wye right legs. These piles of soil were made to create a visual barrier hiding a semi-trailer parking lot. They are subtle terrain profile variations, but they do have quite an impact on the overall perception.

Transloading area completed: lower ground creates a sense of distance on a cramped space

When these things were done, it was time to paint the track. Since we got great results at Ciment Saint-Laurent, we decided to reuse the previous formula: white primer on all track, Krylon Camouflage Brown on the rails and oil paint washes over the ties. If things go smoothly, this task will take about 3 weekends to complete due to drying time and our schedule.

Primed track

Meanwhile, Louis-Marie did a terrific job at modelling Desbiens Street out of cardboard. I'm always surprised how he can be extremely precise with this kind of stuff. When you had the little barn to the scene, it becomes clear this will be a very nice railfanning spot. Having the road slopping gently toward the track makes for a more dynamic and engaging topography.

Desbiens street made out of Strathmore cardboard

Framing the scene with the barn

Desbiens Street: scene composition with the barn and left wye leg

Finally, Jérôme decided to remove a few extra layers of form from the mountain on the peninsula. I kind of like the work on texture I did back a year or two ago, but the mountain shape wasn't very convincing. By removing about 5 inches, it seems proportions are better. Less verticality makes the layout appear far more horizontal and longer. This kind of optic illusions counts a lot when trying to convey the sense of place. I'm well aware many will be kind of upset about us tearing apart the only "finished" scenery on the layout, but I honestly don't care. There is no point in keeping stuff that doesn't fit the level of quality work we can now do. It was our second attempt at serious scenery, it wasn't bad, but we can do better.


And after... nobody was hurt!

At this point, it seems the entire mountain will be remove and new scenery will be done from scratch. Clermont topography is far too different to waste time compromising with half-baked previous efforts. Also, it is more and more evident that some visual key elements of the scene will have to be interpreted quite a bit to be more compelling. An example, I located were the small brook is located on the prototype and it doesn't look good on the layout. Some artistic license will be required to better frame the scene and portrait Clermont in a way that feels right and natural.