Computer graphics vs Reality

That’s interesting, because when I saw Final Fantasy: Spirits Within I was tremendously impressed with the photorealistic characters, but at the same time it felt like there was something “off” about them. Apparently, some people had the same reaction to The Polar Express.

I work in video games, and one question that I’m constantly asked is when will games become “photo realistic”.

The answer, of course is that they already have. Or they will tomorrow. Or in ten years time.

Because it depends on the compexity of the scene. A car is relatively simple to model compared to a human being (and, incidentally, another reason why GT5 looks so good: accurate car models are a bi-product of the motor industry).

Over time increasingly complex scenes can be rendered photo-realistically.

Will there ever be a time when we can render anything photo-realistically, within reason?
Well I’d say yes, with two caveats: Firstly the “within reason” is necessary – even if computers could render anything we could imagine, they still couldn’t render anything. And secondly we would need new input methods – at the moment, most objects in a game are effectively “sculpted” by 3D modellers. Ultimately we need faster ways; 3D scanners and such.