Hi, guys!
I'm trying to have a further understanding in how game's cinematics are made.
Generally speaking, a ps3/xbox 360 in-game character has somewhere between 10.000 to 20.000 polygons and the details are simulated with normal maps, right?
But when you watch a game cutscene (cinematic), where realtime rendering is not an issue, how many polygons are we talking about?
What about the maps (displacement instead of normals)?
Example:
The Batman model in Injustice game vs Blur studio's cinematic.
Replies
I've heard of in game characters going as high as 40,000 tris on current consoles.
But for cinematics there aren't really any rules, displacement maps are common, but really it's mostly about getting the best result as quick as possible. Characters can be millions of tris. A big part of it is how many polygons are practical to work with on their hardware. There isn't limits of what kinds of texture maps are used, multiple textures can be used to simulate subsurface scattering on models. Materials can be procedural and animated in ways not possible to simulate realtime, yet.
Kawaii ne?
I think it depends. That's certainly how the new Mortal Kombat games seem to do it.
Obviously there is advantages using the realtime models, like there was already mention, you can see your own character and it's custom items on it.
But the WORST in-game cutscenes are the crudely drawn "motion comics" used in Marvel Heroes Online. NEVER do that if you have any choice, please?