The future has turned out to be further away than we thought. A few years ago, many people hoped that the next generation of videogame system would finally deliver the holy grail of visual engineering - photorealism. It didn’t happen. Perhaps this has been disappointing; but perhaps, too, we should take stock of visual aesthetics and decide what we have to gain, and to lose, from the prospect of true photorealism.

The most obvious problem is an increasing gulf between the logic of appearance, and the logic of behaviour. If you’re going to raise the retinal stakes to a photorealistic degree, a comparable increase in gameplay coherence will also be necessary. Parts of Gran Turismo 3 look tantalisingly close to the photorealistic ideal, for instance, but this sensuous believability renders the absence of car damage all the more incongruous, even if we know it to be a commercially determined limitation.

This problem becomes more pressing in videogames that encourage exploration of complex environments. Up until now, gameplayers have agreed to accept that if an in-game object is just a clump of flat pixels, then it has no functional use in the game. But imagine a completely photorealistic environment in some future iteration of Biohazard, and it becomes clear that if something looks exactly like a lamp, or a book, and yet we can’t pick it up or read it, then we are going to feel frustrated at the arbitrary limits placed upon our interaction. Shenmue, though nowhere near photorealism, at least makes a creditable attempt at this sort of wildly thing-populated world, but it too has limits: you cannot, for example, simply decide to steal one of the bikes outside the motorcycle shop in Dobuita and go for a ride.

Ideally, a gamer should never arrive at the hard limit of interaction in this way, where a game just tells you: Sorry, no, that’s not part of the game, don’t bother. On the contrary, a gamespace should be like the Einsteinian concept of spacetime. Travel in some lightspeed-defying spaceship to the edge of the universe and you don’t just smash into a cosmic wall; you follow a straight line through space that curves back in on itself and eventually delivers you back to where you came from. This sort of internal coherence has historically been easy to deliver in largely symbolic games - from Pac-Man and Defender to, say, Bangai-O. But in complex, iconic modern games that aspire to visual realism and offer us a host of complex actions, it is far more rare.

The other major problem of taking photorealism as a goal is, of course, the imaginative limitations it imposes. Metropolis Street Racer did an astonishing job of delivering recognisable recreations of central London, for example, but it would be a shame if thousands of still photographs and hundreds of hours of video footage became de rigueur as raw material for all videogames. After all, part of the joy of videogames is the interaction with environments that could not possibly exist in the real world. In this sense, Wipeout Fusion stands as a greater artistic achievement than MSR.

This is not to say that all games should adopt a moody, sci-fi rollercoaster aesthetic. But the distant prize of photorealism does appears to strap blinkers on to some developers while it is still not possible. Here it is the exceptions that demonstrate the rule. Cel-shading’s first appearance in Jet Set Radio was an unforgettable breath of fresh air among the usual slew of pathologically dark, pointy, pseudo-”realistic” videogames. And the work-in-progress shots of Exo’s gorgeously naked cityscape have had observers crossing their fingers in the hope that the developers do not get around to adding textures at all, or at least that they won’t get as texture-happy as so many current designers who smother their spaces in meaningless detail.

One definition of visual aesthetics, after all, might be that it is the art of knowing what to leave out. In the rush for realism in computer graphics, however, it is all too common to throw everything in, and leave the spectator navigating a visual field of hyperspecific effects - excitable lighting, rivets on everything - and unable to see the wood for the trees. Squaresoft’s notoriously expensive CGI-generated Final Fantasy film is a case in point. Watching it, one becomes fascinated by the extraordinarily lifelike movement of a strand of digital hair, for example, or the reflective qualities of an eyeball, in a way that becomes highly distracting.

As a whole, in fact, Final Fantasy stands as something of an exemplary folly in the quasi-art of “realistic” graphics. Not only have the mouths of the actors speaking the lines been digitised and mapped onto their avatars, but every human movement, for however anonymous a gun-toting grunt, has been motion-captured from real humans in ping-pong-ball suits. You get people to do all this - and then you throw the people away. The end result has only a sort of eerie impressiveness, like watching a play performed by ghosts. Why try so hard to replicate natural human movement, when live-action cinema does it so well? The style of “hand-drawn” animation in computer graphics has hardly gone out of fashion since Lara Croft’s first appearance. It would have been more interesting for Squaresoft to have gone down the path trodden by Shrek, whose savage parody of Disney sentiment puts computer animation to purely fabulous uses.

In order to grow into their full artistic potential, videogames will have to embrace other ways of seeing than the photorealistic. But the catch is that it might only truly happen when photorealism actually becomes possible. Once photorealism is no longer the longed-for finish-line of a technological race but a completely routine way of doing things, then we might see, in reaction, a new flowering of more imaginative videogame art. Just as painting was rescued from its cul-de-sac of painstaking realism by the arrival of photography, and subsequently exploded into impressionism, surrealism and abstraction, so the advent of true realism in computer graphics might free us to become ever more unrealistic.

Leave a comment: