📖 Read Next: What even is Synthwave →

📖 Go deeper: Low Poly: A Future we remember from the Past →

🧥 View: The Shinobi Drop →

🎧 Tune in: Blue Sky DnB →

Alter your perspective.

Learning to Speak in Polygons

As developers became more comfortable with 3D, the technology improved and so did their understanding of it.

Compare the earliest PlayStation games with those released near the end of the console's life and the difference can be enormous, despite the machine itself remaining essentially the same.

Developers learned where they could cheat.

They became better at hiding geometry, managing memory, constructing textures, controlling cameras and deciding where to spend detail. Art styles increasingly evolved around what particular hardware did well rather than treating its limitations purely as obstacles.

The next generation pushed that process much further.

The Dreamcast, PlayStation 2, GameCube and Xbox could produce smoother characters, larger environments and more elaborate effects. Games such as Sonic Adventure, Jet Set Radio, Crazy Taxi and The Legend of Zelda: The Wind Waker increasingly demonstrate something that would become obvious over the following decades: greater hardware power doesn't have to mean greater realism.

It can also mean greater control over style.

That distinction becomes important when we look at what happened next.

When Low-Poly Became Old

For a while, early 3D aged awkwardly.

Pixel art had already spent decades developing as an established visual language, and its abstraction could remain remarkably clean when viewed years later.

Early polygonal graphics were different. They belonged to a technology advancing so quickly that games could look dated within a few years.

New consoles advertised smoother models, larger worlds, higher resolutions and increasingly realistic lighting. Jagged polygons and blurry textures were exactly the things developers were trying to leave behind.

By the PlayStation 3 and Xbox 360 era, mainstream games were moving rapidly toward high-definition realism.

Low-poly graphics hadn't disappeared, but the rough look of early console 3D no longer represented the cutting edge.

It represented the past.

Abstraction Didn't Disappear

Simplified 3D hadn't stopped being useful, though. It was finding other forms.

Minecraft is the obvious example.

Its block-based world belongs technically to voxel graphics rather than the polygon-starved appearance of the original PlayStation, but its success demonstrated something important at a time when much of the mainstream industry was pursuing increasingly detailed HD graphics: a modern 3D game didn't need to become more realistic to feel expansive.

In Minecraft, the abstraction is inseparable from the design. The same blocks that represent mountains, forests and caves are also the materials players use to dismantle and reconstruct them. The world's simple visual language makes its underlying system immediately understandable. You don't need a detailed simulation of a tree when the important thing is recognising that it contains wood, knowing that you can break it apart, and imagining what you might build with it.

That simplicity also leaves an enormous amount of room for the player. A handful of blocks can suggest a house, castle, railway or entire city because the game provides enough information for us to understand the structure without needing to represent every detail.

Minecraft wasn't reviving the PlayStation aesthetic. It was doing something arguably more useful to this history: demonstrating that abstraction itself had never become obsolete.

The specific appearance of early 3D had fallen out of fashion.

Eventually, developers would start looking at that again too.

The Indie Revival

Modern low-poly games aren't usually low-poly because the computer can't handle anything else.

That's the fundamental difference.

Indie developers began deliberately returning to the visual language of early 3D, eventually producing games that recreated not only simple geometry and low-resolution textures but some of the quirks developers in the 1990s had worked hard to disguise.

Polygon wobble, texture warping, pixelated resolutions, short draw distances, chunky character models and restricted lighting all became available as deliberate stylistic choices.

Modern engines can correct these things easily. Developers now write shaders to put some of them back.

The Haunted PS1 community provides one of the clearest examples. Its downloadable Demo Disc projects are built for modern computers using engines such as Unity, Unreal and Godot, but deliberately evoke the atmosphere and visual language of late-'90s games. The community itself is broader than strict PlayStation imitation, describing its interest as low-fi and retro art direction rather than requiring every project to reproduce one console perfectly.

That distinction matters.

The revival isn't really about recreating the PlayStation with archaeological accuracy.

Developers are treating early 3D as an artistic vocabulary.

Choosing the Limitation

Modern games also reveal why that vocabulary remains useful.

Crow Country uses deliberately chunky characters reminiscent of the simplified field models of Final Fantasy VII. Creative director Adam Vian has explained that the look partly emerged from building characters out of basic shapes rather than using elaborate modelling tools. He found that the resulting abstraction made the characters appealing, while a camera shader deliberately crushes the image down and introduces distortion. GamesRadar+

For horror developers, ambiguity has another advantage.

A detailed modern monster tells you exactly what you're looking at. A vague figure half-obscured by darkness, low resolution and unstable geometry leaves considerably more unresolved.

Developers behind modern low-poly horror games have spoken about precisely that quality: the abstraction gives the player's imagination space to become involved, while technical-looking restrictions encourage unusual compromises rather than allowing every problem to be solved with additional detail. GamesRadar+

There are practical benefits too.

A small team can produce simplified assets more quickly than a studio attempting photorealism. Stylisation can make assets from different creators feel coherent. Technical requirements can remain modest.

But perhaps the most interesting benefit is simply having a boundary.

Modern game engines offer an intimidating amount of possibility. A developer can choose from physically based materials, global illumination, enormous textures, complex particle systems, procedural geometry and increasingly sophisticated animation.

Having access to everything doesn't automatically tell you what to make.

A restriction does.

If characters can only contain so much detail, you have to decide which details matter.

If the texture is tiny, every mark has to earn its place.

If the player can't see far ahead, you start thinking about what might exist in the fog.

These are old technical problems that have become useful creative questions.

Low-Poly Without the Nostalgia

That is also why low-poly doesn't need to imitate the PlayStation forever.

The wider style now ranges from deliberately crude PS1 horror to clean geometric landscapes, abstract games and highly stylised characters that could never be mistaken for something made in 1997.

What connects them isn't necessarily nostalgia.

It's abstraction.

A low-poly tree doesn't need every branch. A face doesn't need every wrinkle. Water doesn't have to simulate the physics of water convincingly.

The artist decides how little information is necessary before we understand what we're seeing.

Sometimes removing detail can even make a design clearer because there is less visual information competing for attention.

This is something videogames discovered through necessity before artists could comfortably choose it.

When the Flaw Becomes the Feature

There is something unusual about seeing PlayStation-style texture warping reproduced intentionally on a modern computer.

The original developers generally weren't celebrating it. Naughty Dog deliberately avoided heavily texturing Crash partly because the hardware could make those textures stretch and warp. Decades later, another developer might spend time writing a shader specifically to reproduce the effect. All Things Andy Gavin

The technology improved.

Our relationship with the old technology changed.

Once an obsolete limitation no longer has to be endured, it can be separated from the hardware that caused it and considered on its own terms. Polygon jitter becomes movement. Pixelation becomes texture. Low draw distance becomes atmosphere. Simplified geometry becomes abstraction.

Not every technical flaw deserves revival, and limitations didn't automatically make older games better. Developers produced remarkable work because they understood their constraints, not simply because those constraints existed.

That may be the more useful lesson of low-poly design.

Early 3D developers were constantly asking what they could afford to show.

Modern developers can ask a different question:

How much do we actually need?

Early developers spent years finding ways around the limits of their machines. Now that many of those limits have disappeared, some artists are choosing to put them back.

What began as the boundary of what a computer could display has become another tool for deciding what a game should look and feel like.

When Low-Poly Was High Tech

Today, low-poly graphics are an artistic choice.

A developer can open a modern game engine capable of realistic lighting, detailed environments and models containing millions of polygons, then deliberately build a character from a few hundred. Textures can be made pixelated, resolutions lowered and even the visual quirks of thirty-year-old hardware recreated through shaders.

In the mid-1990s, there was nothing deliberate about the basic limitation.

If you wanted to build a real-time 3D world on a home console, you had a small number of polygons, limited memory, low-resolution textures and hardware that was still figuring out how to put all of this together.

The interesting part is what developers did with it.

When 3D Was the New Technology

The arrival of consoles such as the PlayStation, Sega Saturn and Nintendo 64 brought real-time 3D into millions of homes.

For developers, this opened an enormous new design space. Characters could move through environments rather than across them. Cameras could change position. Buildings could have interiors, landscapes could stretch into the distance and games could begin constructing places rather than stages.

But early 3D came with severe restrictions.

Every character and environment had to be assembled from a limited budget of polygons. Textures were small, memory was precious and screen resolutions left little room for fine visual detail.

The original PlayStation had its own distinctive problems. Its hardware lacked some features that later became standard in 3D graphics, contributing to the texture warping, polygon jitter and unstable edges now immediately associated with the console.

At the time, these weren't charming retro effects. Developers generally wanted their games to look as good as possible.

That meant learning how to design around them.

Designing What the Player Could Actually See

Crash Bandicoot is a particularly good example because Naughty Dog has documented just how closely its art was shaped by the PlayStation.

Crash's large head wasn't simply cartoon exaggeration. On a low-resolution television, his facial features needed to remain readable. His black gloves helped distinguish his hands as his arms crossed his orange body, while the lighter chest and spots on his back made his orientation easier to read. Character design was being influenced by a very practical question: which details will survive once this model is reduced to a few dozen pixels on screen? All Things Andy Gavin

Naughty Dog made another unusual decision. Rather than covering Crash in detailed textures, much of his appearance came from coloured polygons. The original model used 512 polygons, with textures reserved for small details such as his spots and shoelaces. This gave the team more geometry to work with while avoiding some of the PlayStation's texture-warping problems. All Things Andy Gavin

Even the game's familiar camera was partly a technical solution.

Instead of allowing the player to look freely in every direction, Crash Bandicoot largely moves the camera along predetermined paths. Naughty Dog could calculate in advance what should be visible from different positions and concentrate the console's limited rendering budget there. The first game was designed around a limit of roughly 800 visible polygons in a given view; adding an enemy or platform could mean simplifying scenery somewhere else. All Things Andy Gavin

The restrictions affected gameplay, character design, level construction and camera movement at the same time.

Low-poly wasn't an art department working with inferior tools.

It was part of the design of the game.

Turning Weakness Into Atmosphere

Different games found very different answers to the same problem.

The original Silent Hill wanted a fully 3D town rather than relying primarily on pre-rendered backgrounds. The PlayStation couldn't comfortably draw a detailed city far into the distance, so Konami restricted what the player could see with darkness and fog.

That technical compromise became one of the game's defining creative decisions.

You couldn't see what was waiting further down the road.

Buildings emerged gradually from the haze. Figures became threatening before you could properly identify them. The town felt larger because so much of it remained obscured.

Members of the development team later explained that fog and darkness allowed them to limit visibility while creating exactly the kind of uncertainty a horror game needed. Director Keiichiro Toyama also emphasised that real-time polygons made the moving camera, darkness and fog possible in ways that pre-rendered environments wouldn't. Silent Hill Memories

The limitation didn't disappear.

It found the right job.

Suggestion Instead of Simulation

That happened throughout early 3D.

Developers had to decide which information mattered enough to survive simplification.

A face might only have enough geometry for a nose, eyes and a rough mouth. A distant building might be little more than a box with windows painted onto it. A forest could use flat planes and textures to suggest foliage that the machine couldn't model individually.

Some games avoided the problem by combining different techniques.

Final Fantasy VII used relatively simple polygonal characters while exploring detailed pre-rendered backgrounds, then switched to different models when battles or cinematics demanded them. Resident Evil similarly placed real-time characters over fixed backgrounds, allowing environments to contain far more visual detail than the PlayStation could have rendered as complete real-time 3D scenes.

Others embraced abstraction more openly.

The important skill was deciding where detail actually mattered.

A recognisable silhouette might be worth more than accurate anatomy. A strong colour difference could communicate more than another hundred polygons. Animation could sell a character that looked crude in a still image.

Players' imaginations handled much of what remained.

The game provided enough information to establish the world; your mind quietly completed it.

• Home

• Blog

• Store

• About