BUILDING THE 3D SCENE

I could have used a stock 3D asset, or paid someone to build the whole scene for me. I didn't, because the entire point of this rebrand was proving I could actually do the work, not just direct it. If I was going to call myself a creative developer, I needed a homepage that backed that up.

I already had a bit of experience with Blender, but I wanted to push past the basics and actually build something ambitious. I started with private lessons with Umberto, a couple hours a week, and just kept showing up. Every shortcut, every tool, every stupid mistake, I learned it by actually building this specific scene instead of following generic tutorials that teach you nothing you'll remember.

It started with a layout, deciding what objects would exist and where. Then came the big shapes first, the desk, the shelving, the walls, blocked out roughly to get proportions right before adding any detail. From there I moved object by object: the guitar, the books, the totem, each one modeled from scratch, checking references constantly so things didn't look off. Once the geometry was done, I moved to materials, PBR textures, UV unwrapping, testing how light hit different surfaces. Metal was the hardest to get right, it took several failed renders before it actually looked like metal instead of grey plastic. Then came lighting the whole room so it felt like one coherent space instead of a pile of separate objects sitting next to each other. Every pass, I'd render, notice what looked wrong, go back in, fix it, and render again.

I had to pick the right technology to get this scene onto the web. Three.js is probably the better long term answer, more control, more power, and it's something I'm actively working toward. But I don't have the skills or the time for that yet, so I went with Spline instead, a tool built exactly for getting real time 3D onto a website without writing a rendering engine from scratch.

This part I didn't do alone. Emilie helped bring the scene over from Blender into Spline, handling the import, fixing what broke along the way, and getting everything set up for the interactions that would come next. Moving a scene between two completely different pieces of software always breaks something, and having someone who actually knew Spline made that transition far smoother than it would have been on my own.

The last step was getting the scene running properly on the actual site. Embedding it cleanly into Webflow, making sure it didn't fight with everything else happening on the homepage. Performance is still a work in progress, Spline scenes are heavy, and I'm actively working through optimization to get load times where they need to be. That's where the scene stopped being a file sitting in Spline and became part of the actual site people visit today, imperfect but real.