Freytra Peaks making of

Hi, I'm Sven, lead artist of SuperTuxKart.

In this blog post, we'll look at the creation process behind the upcoming track "Freytra Peaks", which will be included in the main game with the release of STK Evolution. This new track will retire the venerable "Snow Peak" after many years of faithful service. You can get early access to Freytra Peaks now through our 1.5 donation package.

I started work on this track pretty early, not long after I began "Black Forest". At the time I was still exploring and experimenting with track creation. I've learned a ton since then, but honestly, that makes revisiting this project even more enlightening!

Carving out the Layout

Every track starts with a vision. For Freytra Peaks, I knew I wanted a winter track built around one signature moment: a massive jump that would give players that exhilarating rush of flying through mountain air.

I began by sketching the road layout spontaneously, letting the terrain and that central jump dictate the flow. This intuitive approach can produce natural-feeling tracks quickly, but it requires careful refinement. Early versions included a corkscrew turn near the end that I ultimately cut. It added track length without adding excitement. In kart racing, every second matters. Players should spend their time in memorable moments, not filler sections.

The first road sketch

The refined layout, closer to the final version

The challenge with organic track design is ensuring players get great views of what's coming. When you're racing at speed, sightlines are everything. For Freytra Peaks, the open mountain setting gave me flexibility. I could sculpt the landscape around the road to create eye-catching views and telegraph upcoming turns. A sprawling mountain range as your backdrop is forgiving that way.

Starting to build around the road

Today, my process has evolved. I now map out key visual landmarks first, then build the road to showcase them. But Freytra Peaks taught me those lessons, and the final result captures exactly what I was after: the feeling of racing through isolated mountains, with nothing but snow, stone, and sky around you.

Painting the landscape

With the road layout established, it was time to build the world around it. My goal was clear: create the atmosphere of a remote mountain refuge, isolated peaks, a tiny village clinging to the slopes, and vast stretches of untouched snow.

I started with placeholder assets, existing trees, houses, rocks, and background mountains from our library. This is a crucial step in track development. Placeholders let you evaluate composition, scale, and atmosphere quickly before committing to custom modeling. You can iterate on the overall feel without getting bogged down in details too early.

The terrain itself required careful sculpting to achieve those characteristic alpine forms, sharp ridges, sweeping bowls, and dramatic elevation changes. For the distant mountain ranges, I used Blender's A.N.T. Landscape addon to generate realistic terrain, then created height-based masks to blend snow and exposed rock textures naturally. Blender's texture baking feature let me convert these procedural materials into optimized textures that work seamlessly in Antarctica, our game engine.

Generating mountains

The background mountains don't need to be highly detailed, their job is to create depth and sell the scale of the environment. They do that job perfectly, framing the track in an expansive alpine wilderness.

One technique I'm particularly fond of: the snow layer. By duplicating the landscape geometry, raising it slightly, and tucking the edges under the road and rocks, I created a convincing blanket of snow that gives the environment real depth. It's a simple trick, but very effective.

Adding the snow layer

Driving into the snow

The Devil's in the Details

With the landscape established, I began populating the world with the details that would bring it to life. This phase is where tracks truly find their character and where the bulk of development time goes.

I created custom props for the track: signs, barriers, small structures, and environmental details. Most of these initial models served the track well, though I knew the main lodge building would need a complete overhaul eventually. Sometimes you need to see an asset in context before you know what it really needs to be.

Various objects created for Freytra Peaks

At this point, the track had its bones. I set it aside to work on other projects, knowing I'd return with fresh eyes later.

From Snow Peak to Freytra Peaks

When I returned to Freytra Peaks years later with STK Evolution on the horizon, I saw it with new eyes and new standards. The trees and buildings, in particular, needed to match the quality bar we're setting for Evolution.

A landscape from Snow Peak, the old track replaced by Freytra Peaks

A landscape from Freytra Peaks

The trees presented an interesting challenge. I adapted three tree models I'd created for another track, modifying their textures to carry snow. The trick was making snow look like it's actually resting on the branches, not just painted on. Pure white snow looked flat and unconvincing. The solution: blending the original green foliage with snow texture, letting some green peek through while the snow extends slightly beyond the branch geometry. It creates that heavy, winter-laden look that sells the environment. Given unlimited time, I'd love to add actual 3D snow clumps on the branches, but the current approach strikes a good balance between visual quality and performance.

The trees, with and without snow on their branches

For the village buildings, I crafted one detailed house, then created variants by modifying it. A practical approach that maintains visual consistency while populating the scene efficiently. Smart asset reuse is part of professional game development.

The village buildings

The track's signature jump created an interesting problem for reverse mode, players couldn't make it backwards. I considered building a reverse jump, but the terrain didn't support it elegantly. Instead, I carved three narrow paths through the forest that feel like mountain hiking trails. It's a solution that fits the environment naturally while solving the gameplay challenge.

Before releasing the track in the donation package, I rebuilt the main wooden lodge from scratch. I also refined countless smaller elements: icicles clinging to rocky overhangs, smoother texture transitions across the terrain, more detailed rock formations. These touches might seem minor individually, but collectively they create polish and immersion.

The new wooden lodge

What's Next for Freytra Peaks

Ask any artist when their work is truly finished, and an honest answer will often be: never. Given unlimited time, I could refine Freytra Peaks endlessly, tweaking textures, adjusting geometry, adding new details. It's the George Lucas problem: without discipline, you end up releasing Special Editions every few years with "enhanced" Jabba scenes nobody asked for.

Before STK Evolution's full release, I'm planning a final polish pass: more sculptural roadside rocks, a more dramatic start/finish area, improvements to the main jump, and enhanced tunnel atmosphere ; but for the most part, I'm calling Freytra Peaks done, focusing my attention on other tracks.

Indeed, beyond Freytra Peaks, I'm bringing this same quality standard across the entire game. "Candela City" and "Old Mine" are currently on the workbench, getting the Evolution treatment. Every classic track deserves to meet the bar we're setting.

Join the Journey

Inspired to create your own track? The SuperTuxKart community welcomes creators of all skill levels. Check out our track creation guide at supertuxkart.net/Making_Tracks to get started, maybe your creation will feature alongside Freytra Peaks someday!

I'd also like to thank Samuncle and Alayan for their help in drafting this blog post.

My Return to SuperTuxKart

In 2017, I left SuperTuxKart on the highest note imaginable - a standing ovation at the Blender Conference in Amsterdam as we launched the mobile version. I believed then, and still do, that it's better to end with a firework than fade slowly to irrelevance. I thought that chapter was closed. But sometimes, a project you love calls you back - not because your story wasn't complete, but because there's a new story to write.

The team asked me to return not because I have all the answers, but because I remember the questions we used to ask ourselves during those seven years, from 2010 to 2017.

I am Samuncle, former lead artist of SuperTuxKart.

Candela - where we proved open-source could be magic

Humble Beginning and Angry Letter

My journey to SuperTuxKart began in 2007 with youthful exhuberance, ambition and questionable priorities: I wanted to hack my school's Wi-Fi, which apparently required Linux. Installing GPU drivers was a nightmare back then, and the common wisdom was "download a game to test if it works." That game was SuperTuxKart.

But the seeds were planted much earlier. Since playing DOOM on DOS machines as a child, I'd been obsessed with making games. I remember cutting out cardboard pieces, trying to construct my own "3D" worlds. When Far Cry arrived, and later Crysis, something crystallized in my mind. Not the gameplay itself, but the mythology of it - a game so technically advanced it became a benchmark, a statement of what was possible.

I wanted to create that. Not necessarily an FPS, but something that carried that same audacity.

In 2010, I read about a new SuperTuxKart beta release and downloaded it. I was devastated by what I saw. The graphics were… not what I'd hoped for. So I did what any passionate teenager would do: I sent an angry letter to the developers telling them they were bad at their jobs.

One of the artists responded with a challenge I didn't expect: "Do better."

I left for a week. I worked on something called Hacienda. When I returned, I had created something better than anything in the game at that time.

That's how I joined SuperTuxKart, part reckless, part intimidated, wholly sincere. I literally started by complaining. But beneath the bravado, I was scared and unsure. I was an apprentice who didn't yet know what magic looked like. What followed was a slow process of refining every map, learning from existing contributors, and gradually discovering what I could contribute.

Interestingly, I wrote to several other open-source projects around that time (less harshly, I promise). They all rejected me. So I stayed with SuperTuxKart. Sometimes the best path chooses you.

The War for Better Graphics

I quickly evolved from angry teenager to lead artist. I could see that many open-source games were failing not in code, but in visual ambition. So I invested everything into raising the art quality - from the beautiful shores of Gran Paradiso Island to the lost ruins of Cocoa Temple.

But art alone wasn't enough. We were constrained by Irrlicht, our rendering engine, which had severe limitations you wouldn’t even imagine in 2026. For instance you couldn't use more than four textures per object. Linux distribution policies made it worse - they forced us to use vanilla Irrlicht, rebuilding our game and discarding our modifications. We were trying to reach for the stars while wearing chains.

So we made an audacious decision: we forked Irrlicht entirely and built Antarctica, our own engine. I pushed relentlessly for this. My collaborators, Auria, Hiker, and later Benau, all extraordinarily competent and professional - somehow tolerated my increasingly wild demands. I'm grateful for their guidance and patience as I learned what leadership actually meant.

For me, the release of Antarctica was the realization of that childhood dream. Crysis wasn't about first-person shooters - it was about philosophy. It was about creating something so technically and artistically aspiring that people asked, "How is this even possible?" I wanted SuperTuxKart to ask that same question for open-source gaming. Not 'good enough for FOSS.' Not 'impressive given the constraints.' Just… what if we refused to accept limitations at all?

It took years, but by 2016 we were making it real. Professional trailers, bold marketing. We were making people ask, "Wait, this is Free and Open Source?!"

There's an uncomfortable truth about why SuperTuxKart succeeded where many open-source games struggled: I never looked at other open-source projects as my measuring stick.

I studied Pixar. Disney. Blizzard cinematics. I asked: 'What makes people feel something when they see this?' Not 'What's achievable with volunteer developers?' but 'What's worth achieving?'

This wasn't arrogance, it was survival. If players compared SuperTuxKart to other FOSS games, we'd be niche forever. But if we could make them forget it was open-source, if the first reaction was 'Wait, this is beautiful,' then we could talk about freedom, and community, and all the values that matter.

By 2017, everything had aligned. We had our engine. We had stunning tracks. We were launching on mobile. Standing in Amsterdam, receiving that applause, I felt complete.

But I also had a new job. Initially, I was naive enough to think I could do both but the 9 to 6 and sometime 7 or even 8 made it impossible, so I left. I spent the years that followed on diverse work: 3D artist, software engineer designing rendering engines. Ironically, I ended up on the other side of the table, working on technical problems similar to what I'd tackled at SuperTuxKart. The work was excellent, but less artistic, more technical. I missed art for art's sake. I missed the permission to be unreasonable in pursuit of beauty.

And I missed SuperTuxKart. I'd known it the moment I stepped off that Amsterdam stage. I knew that chapter was closed, that I'd never contribute that way again.

Until today.

Why Now?

When reading the Evolution roadmap and connecting with the new team, I realized they were asking the same question we obsessed over almost a decade ago: "What can open-source gaming actually become?"

Today, SuperTuxKart has grown tremendously - more players, network mode, ongoing development. The team reached out because they want to recapture that mindset: to redefine what open-source games can be, standing on the shoulders of what we built together. Looking back, maybe we were naive. But that naivety gave us permission to attempt the impossible.

Guiding the Next Generation

I'm particularly excited to work as the executive producer, helping our team leader Alayan and our new lead artist Sven. Sven has an incredible talent, and the same energy I had at his age. My role isn't to impose my vision - it's to help them find theirs, and give them permission to be uncomfortably ambitious.

We stand on the shoulders of giants - Auria, Hiker, and countless contributors who believed that open-source deserved excellence, not excuses. Now it's our turn to be those shoulders for the next generation.

Sometimes the most valuable thing an old wizard can offer isn't answers, but the courage to ask: "Why shouldn't we?"

Let's see what Act II brings.

Sam