That's probably a friendly robot. All robots are friendly, right? Asimov said so.
We catch our first glimpse of the “Death Bog” glimmering miragelike in the distance… and our first glimpse of one of its denizens up close and personal. When I first wrote this chapter, the “Death Bog” wasn't a swampy artificial nature realm, but instead it was going to be a “Death Labyrinth,” a techno maze area with lots of cool death traps. But the test renders I did were less than stellar, and then I decided I'd riff off of the “human hunting” idea introduced in the game show sequence earlier in the chapter. So, a few quick dialogue changes later, the Death Bog was born.
This isn't really a “science blurb;” it's just me reflecting on the current state of 3d graphics. Recently, Unreal released a sneak peak at the latest version of their Unreal gaming engine and it's more than just an incremental step forward in real time 3d rendering… but I'll get back to that in a minute. A thing that occurred to me a while ago is that 3d art and animation appear to be coming full circle back to the real thing. What I mean by that, is that in a lot of the big budget movies you see these days, and the blockbuster video games as well, many of the 3d props and characters you see weren't modeled by some 3d artist at a workstation; they were scanned from a real world actor or a real world object. Most of the character animations and facial expressions aren't created by an animator; they're motion captured directly from the actor reading the lines and from stunt people performing the stunts. And yes, I know, it still takes a whole team of artists to refine all of that and integrate it into the final product, but the point is that production level 3d art is increasingly less a stylized craft made to mimic life, and more the actual projection of real life into a virtual space.
What the new Unreal Engine is capable of is just mind blowing. In the past, a 3d artist had to take an original scan – which consists of gazillions of polygons – and basically sculpt a lower resolution mesh over it and use some clever material tricks to make the simplified model look more detailed than it actually is. That's because it would take an obscene amount of computing power to calculate simulations using those original gazillion-polygon scans. But with the new Unreal Engine, they can pretty much take those original scans, without bothering to simplify their geometry, and throw them right into the mix with no appreciable loss of performance. I'll admit, I'm more than a little concerned about how much disk space all those crazy dense models are going to take up, but I still have to marvel at the fact that you can even run something like this on your home computer or next gen gaming console without it crashing or catching fire.
So why am I blurbing about this? No reason really; I just thought it was interesting. I had a disastrous flirtation with Unreal Engine earlier this year and have been itching to have another go at it… which I certainly will after I finish this chapter, hopefully in the next couple of months. I feel like it's the natural progression from what I'm currently doing to what I hope to be doing in the near future.
Comments
Please login to comment.
Login or Register${ comment.author }} at
${ comment.author }} at