(This is Part Two of an article about the evolution of computer games into a form of virtual reality, largely as witnessed by the fanatic computer gamer who writes this blog. If you’ve already read Part One, you might want to skim over it again, because I’ve added several paragraphs of material.)

In early 1997, a revolution occurred in computer gaming on DOS and Windows PCs.

I was lucky enough to be in a position to witness this revolution from the very beginning, something most gamers weren’t. In January of 1997, I was working on a project for a local software company and to help me accomplish it they loaned me a computer that was considerably better than the aging DOS/Windows machine I already owned. The loaner had a 200-mhz, first-generation Pentium processor (still more-or-less state-of-the-art at the time and vastly superior to the aging 66-mhz 486 machine I used for word processing and gaming) and something else that was on the bleeding edge of microcomputer technology: a 3DFX Voodoo graphics board.

For much of that decade, computer equipment manufacturers had been attempting to create 3D accelerator cards for microcomputers and gaming consoles that would allow programmers — mostly game developers — to write code for three-dimensional animation that would move polygons across the video display of a microcomputer at considerably faster rates than were possible using purely software-based 3D programming, the type that I’d written about in my book Flights of Fantasy. The first Playstation used one of these and its graphics were very good for the time (about 1996), if not quite revolutionary, though they were good enough to turn Lara Croft into the world’s first video game sex symbol (unless you count Zelda and Princess Peach). These graphic cards took over many of the more time-consuming tasks that the computer’s main processor needed to perform in drawing 3D images, freeing up the CPU to handle other tasks while the graphic card did most of the drawing.

Tomb Raider 1

Lara Croft on Playstation 1: The world’s first 3D-accelerated sex symbol

Unfortunately, nobody had been able to produce a 3D accelerator card that had all the capabilities programmers really needed for truly realistic graphics, until a previously unknown company called 3DFX introduced the Voodoo Graphics chip set. A graphics accelerator board built around this chip set could not only draw 3D polygons on a computer display at unheard-of speeds, it could do so in resolutions of at least 640×480 pixels (considered fairly high-resolution on 14″ monitors) with more than 32,000 different colors on the screen simultaneously, an astonishing number when you consider that most computer games of the period ran in 320×200 mode and only put 256 colors on the screen at one time. 3DFX also released an API — application programming interface — for the chip set called Glide that simplified the task of writing programs that used the chips’ capabilities.

The first game company to recognize the potential of the 3DFX chip set and make a game available that utilized it was, not surprisingly, Id Software, the company that had revolutionized 3D gaming and invented the first person shooter with Wolfenstein 3D and Doom. In 1996, Id had released Quake,  a game that I had played and enjoyed but that hadn’t struck me as the kind of quantum leap beyond Doom that Doom had been beyond Wolfenstein 3D. What was clever about Quake was buried deep inside its code, where the average player couldn’t see it. Id’s head programmer, John Carmack, had designed the Quake graphic image so that it could easily take advantage of graphic hardware quite different from the graphics cards that it was initially intended to run on. Just before I received my loaner computer from a local software company, Carmack had released a modification to Quake that anyone could download freely from Id’s Web site. It allowed Quake to generate its 3D imagery using the 3DFX chip set, eliminating the need for much of the 3D code built in to the game program. Being the first gamer on my block — and probably one of the first gamers in the state of Maryland — who had a computer at his disposal with a 3DFX Voodoo board installed, I downloaded it immediately and ran the accelerated version of Quake. The results were breathtaking.

Here’s a detail of the Quake screen running in 320×200 pixel mode with 256 colors on-screen:

Quake without 3D acceleration

Quake running in low resolution in 256 colors

And here’s a detail of the Quake screen in higher resolution with more than 65,000 colors on screen at one time:

Quake with 3D acceleration

Quake running in high resolution with more than 65,000 colors

I’ve blown up these crops to make the differences more obvious, which causes them both to look blurry, but the differences should indeed be clear. And though few gamers had 3DFX accelerator boards in January of 1997, a lot of game developers did, and when they saw what the new boards did for Quake, the revolution began. Everybody who wanted to write 3D games wanted to write them for 3DFX-based graphics boards.

To get a better idea of what happened when 3D graphics were introduced, watch this YouTube video:

Graphical Evolution of First-Person Shooters: 1992-2012

Be sure you’ve got your YouTube settings at 720 or higher when you start watching. The 3D accelerator shift becomes most apparent here with Half Life in 1998 rather than with Quake 1 or Quake 2 in the previous two years, but that may have as much to do with the resolution settings or video capture software used by the creator. It was the accelerated version of Quake 1 that really set off the revolution. But it’s still obvious in this video that something dramatic happened between roughly 1996 and 1998 and that something was 3DFX.

Within a few years 3DFX had serious competition from chip sets developed by NVidia and ATI. By 2000 the company went bankrupt and sold off its patents to NVidia, which along with ATI remains one of the two major producers of 3D graphics boards for gamers. But by then, 3D-accelerator technology had become ubiquitous. Even ordinary computers, those used exclusively for spreadsheets or Web surfing, had 3D acceleration built in, even though in many cases it was never used.

Once the revolution began, 3D computer animation improved so rapidly that by 2004 games like Half Life 2 could produce scenes that looked like they’d been shot in the real world by video cameras:

Half Life 2

Half Life 2: Is it real or is it a video game?

By 2006, games like The Elder Scrolls: Oblivion had even achieved my holy grail of computer graphics realism: You could see individual blades of grass blowing in the wind:

The Elder Scrolls: Oblivion

Blades of grass seemingly blowing in the wind

This was actually a bit of a cheat. You weren’t seeing individual blades of grass. You were seeing transparent polygons with clusters of grass “painted” on them through a technique known as texture mapping. But it looked real and in the world of computer game virtual reality, that’s what counts.

Only a few years over schedule, computer graphics had achieved the vision I’d had for them in 1981, when I’d looked at a picture of a crudely drawn warrior facing down a crudely drawn monster in a very crudely drawn maze on the box for the game Morloc’s Tower:

Morloc's Tower

A crudely drawn warrior in a crudely drawn maze. Computer game virtual reality had come a long way in 25 years.

There was only one element missing from my original vision of computer games as virtual reality: artificial intelligence. MMOs — massively multiplayer online games — allowed real human beings to interact as computer drawn characters, but not really in a way that produced a story. I wanted computer-generated characters who were as artificially smart as they were artificially realistic, so that I could enter a world and create my own story from existing elements, stepping into a kingdom on the verge of war or a conflict between mobsters and becoming involved with the people involved in such a way that I would actually affect the direction of that world’s history.

Computer game designers were working on this too, but without true AI creative substitutes had to be found that would give the player the illusion that they were in a real world interacting with real people. Essentially, designers have gone in two different directions to create this illusion: allowing players maximum freedom within a richly imagined world or forcing players down a pre-conceived story path that only offers a veneer of true freedom. In the next (and, I hope, final) installment of this post, I’ll discuss three games that take very different approaches to creating stories within virtual worlds: The Elder Scrolls: Skyrim from Bethesda Game Studios, The Walking Dead Season One from TellTale Games and Dishonored from Arkane Studios.