I’ve long been an advocate that what drives a virtual world isn’t how real it looks - it’s what you can do, who you meet, and how robust the tools are for collaboration and generating content. Real is for games - you know, the code heads who drool over how brilliant the water is, say, but neglect to mention that the game mechanics leave you, well, wanting more.

But it’s hard to ignore that that the way chips are headed (you’ve heard of the Singularity, right?) and continual improvements in how stuff is moved around networks mean that virtual worlds are headed at least in the direction of having the ABILITY to deliver more realistic environments.

Blue Mars
At the Virtual Worlds conference in LA I had a chance to get more information on Blue Mars, about which I’ve blogged before here and here. Now - it’s vapour until it’s in alpha or closed Beta, but it sure looks impressive. Their plans are for sharded worlds (with the ability to communicate and locate friends on other shards) that have a capacity for up to 10,000 avatars (and those 10,000 can be in the same location, according to the Blue Mars folks). They have a convoluted business model that now INCLUDES user-generated content, about which more in another post, but what defines the world is still the richness of how it LOOKS.

Seeing Blue Mars “live” was a pretty stunning insight into where the eye-candy is headed - WHEN they actually launch is another question as is whether the business models will make sense, and how they deal with bandwidth issues….it may not be so much an issue of whether the technology CAN do it, it will be a question of how to optimize it (or not) for the masses.

Real-Time Raytracing
Meanwhile, IBM, which seems to have their hands in every pie these days, especially when it comes to 3D content, has loaned some of its new toys out for a test drive. In this case real-time ray tracing. And as someone who’s dabbled in Maya and sat waiting while my so-called powerhouse machine does that mental ray thing, it’s absolutely astonishing to me if what I’m seeing is what they say it is: real-time ray tracing, what SHOULD take hours per frame maybe, being done, well…you get the idea…it’s in real time:

Otoy
Virtual world? No. But I’ll call it “deep renders in real time” - and if you can do that server side, then maybe Otoy’s technology isn’t too far off the beam (whether it’s really Otoy that delivers this is an open question).

There’s still a debate over whether Otoy is real or demo… but when you start piecing together the capacity of the new chips and graphic cards you can see where we’re headed.

Second Life
Second Life made the first move from a fairly “flat” render to something a little richer with the introduction of Windlight. I think it’s still an open question what the impact is of graphic improvements like that on user attraction and retention - if virtual worlds are still in the early adoption phase, maybe we shouldn’t be worried so much about cutting off the low-end machines.

But SL has a few problems: one is that the world streams in real time. Blue Mars, as I understand it, will “load” for example before your avatar fully appears in the environment. Whether this means you’ll need to sit and wait a year before the contents have been packed to your PC I’m not sure, they were a bit evasive on the topic. But it implies that the content may not be dynamic.

But even in SL, parallel ‘renders’ of the world are already possible through things like Kirsten’s “shadow viewer” (she launched the latest version this week).

Even shiny bump maps are possible in SL:

But you know…at the end of the day, it won’t take rich forests or rippling water to make a world. These things will come, they’ll attract new users, but it will still be the content and the people that keep them logged on. I’ll go where Starax goes - just lead the way.