Half Life port brings Black Mesa to Sony’s first console

Half Life on the original PlayStation sounds like the kind of project that should fall apart almost immediately. The hardware is too limited, the game is too complex, and the gap between late-1990s PC design and Sony’s first console is huge. Yet it’s working. An early version of a Half Life PS1 port is now running on real hardware. The build already features playable levels, enemies, weapons, animations and analogue controls. It’s rough, of course. Performance sits at around 10 to 15 frames per second, with the developer aiming for 20 fps as optimisation continues. There are bugs. Visual problems too. But this isn’t just a tech demo that lets you walk around one small room. The goal is much bigger: a complete version of Half-Life, rebuilt to work within the limits of the original PlayStation.

The engine has been built specifically for PS1

This port doesn’t rely on an existing PlayStation project or a simple conversion layer. The engine has been written from scratch for the console, while recreating the systems Half-Life needs from Valve’s GoldSource engine. That’s where the real difficulty starts.

Half Life isn’t only about drawing corridors and letting the player shoot enemies. The game relies on scripted scenes, lifts, doors, trains, level transitions, enemy behaviour, animation events and carefully timed sequences. If those systems don’t work, Black Mesa quickly becomes a collection of empty rooms.

The project uses PSoXide, a Rust-based PlayStation development suite that combines an emulator, software development kit, engine and editor. That setup gives the developer direct control over the whole pipeline. Bugs can be tested, measured and fixed without jumping between a pile of disconnected tools. For a project this demanding, that matters a lot.

Memory is the biggest problem

The original PlayStation has very little RAM. Half Life, meanwhile, was built for PCs with far more memory available. Models, textures, sound, dialogue, weapons and enemy data all compete for space. You can’t load everything at once. Not even close.

The current solution streams assets from the CD depending on the map being played. It also uses Half-Life’s visibility data to avoid processing areas the player can’t actually see. That reduces the amount of geometry the console needs to handle at any one moment. The build tools also check memory use for every map and every level transition before creating the final disc image. That helps catch problems before they appear on real hardware.

So far, all 96 single-player maps fit within the available memory budget, along with the weapons and enemies they require. That doesn’t mean the whole game is finished or stable, but it does show that the full campaign may be possible. That’s a major step.

The graphics need careful compromises

One of the most obvious problems comes from the PlayStation’s affine texture mapping. Put simply, textures can wobble and warp as surfaces move closer to the camera. Anyone who has played an original PS1 game will recognise the effect immediately. The port tackles this by splitting large triangles into smaller ones. The asset tools also choose triangle diagonals that reduce distortion and repair visual gaps caused by the extra geometry.

There’s a catch, though. More triangles can make textures look steadier, but they also increase the load on the CPU, GPU and Geometry Transformation Engine. Split too much geometry and the frame rate drops. Split too little and the image starts to bend.

So the developer has to tune each area carefully. It’s a constant trade-off between cleaner visuals and playable performance. The finished port won’t look identical to the PC game. That isn’t realistic. The resolution will be lower, textures will move, detail will be reduced and performance may vary from scene to scene. The more useful question is whether it still feels like Half-Life.

Can the combat hold together? Do the scripted moments work? Does moving through Black Mesa still have the same pace and structure? Those are the details that will decide whether this becomes a playable conversion rather than an impressive experiment.

Players will need their own Half-Life files

The project is expected to become open source once it reaches a more complete state, but there are no plans to release a ready-made disc image containing Valve’s original game data. Instead, players will need an installed Steam copy of Half-Life. A converter will then use those files to build a PlayStation-compatible image. Rust will also be required to run the tool.

That approach keeps the engine project separate from the commercial assets. It also means the conversion process becomes part of the release, rather than simply distributing a modified copy of the game. It’s a practical solution.

This is already more than a proof of concept

The Half-Life PS1 port still needs a lot of work. Performance must improve, visual bugs need fixing and the full game has to be tested from start to finish. Still, the project has already crossed an important line.

Half-Life is running on real PlayStation hardware with working enemies, weapons, controls and levels. The engine streams data, manages memory and handles systems that the console was never designed to support at this scale. That doesn’t guarantee a finished release. But it does prove the idea works.

Spread the love