It sounds optimistic, I know. Sculpting a 3d version of the Matterhorn, Europe’s most iconic mountain, based on a few photos. Someone with more experience of 3d sculpting would probably have laughed and told me to make a wineglass instead. Since nobody did, I gave it a go.
I picked up the basics in working with terrains in Blender from David Ward’s excellent CG Cookie video series on Creating an Island Environment in Blender. Wanting to take if further, I figured I’d have a play around and make a mountain of my own. And when it comes to mountains, I’m probably not alone in thinking of the Matterhorn with its distinctive triangular profile.
Following the steps in the earlier part of the tutorial and pushing and pulling some vertices around for a while got me to an approximate shape. The rest was done in sculpt mode with dynamic topology enabled. I found it easier (and more fun) to sculpt once I’d applied a matcap shader. It felt much more like painting then.
You can see the shader I prefer in the screenshots. If you haven’t applied matcaps before, you can find them under Display (first press N if your menu isn’t showing) where you can preview or select from a number of different matcap choices.
Aside from in the character creation tutorial, I haven’t done a whole lot of sculpting in Blender. It was frustrating at first but the more I learned, the more I began to enjoy it. For my project, Grab and Scrape were the two brushes I used most, with Pinch, Crease, Clay Strips and Flatten all coming in useful.
Running into Difficulties (More Italian Photographers Needed)
The challenge was in getting the dimensions of this very large, organic object to look right when recreated in 3 dimensions on my screen. I had found some photos to use as background guides, which helped. However, this was also the source of my problems, it turned out…
If you’ve spent any amount of time photographing and generally gazing in awe at the Matterhorn, as I have, you might have the crazy idea that you know it. Good quality photo references will reinforce that feeling.
Actually, there’s not one but two problems with this scenario. I’ll talk about the second one further down. The first problem was something I ran into while working on my model. Everyone takes photos of the Matterhorn from its photogenic sides. These are the north face, as seen from Zermatt, and even more frequently from the east side, as seen from the direction of The Gornergrat (with extra points if you get the picturesque Riffelsee lake in the foreground). Images of Il Cervino from the Italian side are few and far between, and I couldn’t find anything from the west. So it was fairly easy to get something that looked convincing from one angle, but much more difficult to figure out how the ‘back’ (sorry Italy) of the mountain looked.
Let’s be honest, everyone loves the triangular eastern face. The other aspects, not so much. This resulted in a very uneven approach, with one and a half sides of a 3d object looking not bad at all, and the poor, neglected-by-photographers remainder of it being guesswork. I tried, I really did, but my version of the south face wasn’t anything that you’d want gracing your postcards.
As for the west, I discovered that when you model a ridge following references from several different angles, you can miss the fact that it is a single ridge. You then find yourself in the embarrassing situation of a double ridge, each looking something like the genuine article from a certain perspective, but together creating an entirely new geographic feature that should never exist. Of course, an aerial view would have been helpful but at that stage I hadn’t been able to locate one.
Also, with different photo references showing light falling on different parts of the mountain, certain features appeared flattened and others were thrown into contrast. This made it easy to make the mistake of over-sculpting some parts of the rock face. It was only late in the process that I discovered that old drawings of the peak were a far more reliable record of its surface features. Again, though, the artists of old had the same preference for the Zermatt side.
So yes, while it’s an interesting experiment to try, I learned that creating an organic 3d model with photo references that don’t cover all aspects, and which has an irregular shape, is unlikely to be a huge success. Still, I like a challenge and it did give me a chance to get comfortable with some of Blender’s sculpting tools.
Taking a New Perspective on a Familiar Feature
Problem number two only became apparent when at last I discovered a way to get an accurate model of the Matterhorn’s topography into Blender. If only I’d had that to guide me from the start! However, it was interesting to put both models in the same scene and see how my mixture of photo-referenced sculpting and guesswork compared with the real thing.
This was where I discovered that the ground-level view of a human pointing a camera at a very tall object is not the most reliable one for a 3d model reference. That might sound obvious. Maybe it should have been obvious to me while I was working on this.
You see, the iconic image that we know and love comes from our ground-level perspective. Even if every would-be artist learns about foreshortening at some point, it’s easy to forget when every image you’ve seen of something, paired with your own memories of it, come from being down low and looking up at at it, high and distant. That view becomes the ‘real’ view to you. I found it surprising when I finally saw the Matterhorn’s peak the way you’d see it from a helicopter. It’s skinnier than you’d expect, with more extreme angles. Go up further, until you are directly above, and the mountain is almost like a letter x with four sharply creased ridges.
Despite some obvious flaws and the whole foreshortening issue, I’m not disappointed with how my version of the Matterhorn in 3d turned out. Actually, I kind of like it. It looks familiar. It’s the Matterhorn as I remember it, the way it’s stored in my memory (flaws and all), and really isn’t art our view of things rather than a perfect recreation of them?
Most of my efforts went into the peak and I knew my model would lack any accurate representation of how the land lies lower down. As you’ve probably gathered from this post, it doesn’t look good from every angle either. But if I put both models in the same scene, to a similar scale, they match up pretty well in places. And with all the challenges involved, this was a pleasant surprise. I’ve decided ‘in places’ is plenty good enough.
Want to see a wonderful creative tribute to the Matterhorn? To celebrate the 150th anniversary of its first successful ascent on 14th July 2014, talented Zermatt-based artist Amanda Schmid is working on a series of 150 paintings and artworks of the Matterhorn. Along with the time she has devoted to painting it over the years, lucky Amanda lives with a view of this peak and it’s hard to imagine anyone more familiar with it.
Not only are her resulting works beautiful and varied, but they are also have an instinctive accuracy that tells you that the Matterhorn is an old friend of hers.
Visit Amanda and see her growing gallery of Matterhorns in different media at 150 Matterhorns.