Vigormortis said:
RhombusHatesYou said:
Lemme guess... upscale version of mocap/motrack visualisation systems? Cool tech but not relevent. You use down scaled models and BG that a real time rendering engine can handle with rough approximations of lighting. It's about as much a 'certain degree' of 'avatar calibre graphics' as a pushbike is a 'certain degree' of car. The biggest barrier to realtime photorealistic graphics tech is the complexity of calculations required for realistic behaving light, the complexity of which increases semi-exponentially (it's kind of hard to explain exactly how it works without giving a crash course in CGI lighting) for each polygon added to the scene.
Of course, you can argue that to the layman the criteria for being 'avatar calibre graphics' is much lower... and seeing as many would probably have difficulty telling the difference between 'avatar calibre graphics' and high end PC game graphics makes their criteria pretty much useless and the claim of 'avatar calibre graphics' the marketing bullshit of a knobber who thinks 'High End PC graphics' wouldn't be enough of a drawcard.
Um...no offense but you seem to be taking my comments both WAY too personally and not for what they are. Not sure if it was your intention, but your explanations comes off as a tad condescending. Just saying.
I'm not at ALL saying Microsoft's supposed claims are possible. Not even the slightest.
I am well aware of the "complexity" in rendering detailed CG imagery. In real-time or otherwise. All I was saying in my previous post was that rendering highly detailed CG video in real-time
is possible. The cameras used on the set of Avatar did just that. It's just that rendering such imagery with a level of detail even a quarter as detailed as was seen in the final cut of Avatar is impossible. (with todays tech)
Read his post again.
The CGI rendered in real-time for Avatar was
not highly detailed. That was the whole point of his argument. The CGI used on set at Avatar was basic and undetailed. It provided Cameron and co with a working skeleton of how everything would be positioned,
and nothing more. The camera's on Avatar rendered a very basic skeleton of the world, which the animation and rendering then went over and overhauled in practically every way imaginable.
It's no different to when Jackson and Weta used mo-cap to create Gollum for Lord Of The Rings. When Andy Serkis performed his lines and actions, there was
a very, very basic blocky model of Gollum that acted out the moves. It still took hundreds of hours of fine-tuning, extra animation and rendering to create Gollum as he was seen on screen.
Commercially available real-time CG is no more advanced than the latest high end PC games. That's because videogames are now the main area of focus in rendering computer-generated imagery in real time. There's a reason people were blown away by Crysis back when it first came out. It wasn't impressive just for a game, it was impressive by CG standards full stop.