Since I am new to all this (modelling, animating etc.) I was wondering if someone could enlighten me, or point me as to where I can find out about the following. How are scenes in films such as Elephants Dream and other CGI masterpieces like The Incredibles made? Looking at some of the production files for Elephants Dream and from stuff I’ve seen in the past, each scene in a CGI movie is made up of multiple animated scenes, all rendered separately (so in ED for eg this would be the two characters walking along, then the background would be rendered separately). How are these two parts then meshed together and how can it be done so they look like they are one.
Basically, can someone give me an overview of how CGI films are made?
To “mesh” the scenes together a video NLE (Non Linear Editor) program is typically used. Blender has one built in, the Video Sequence Editor. Most / (many?) scenes can be split whenever the camera changes, although a camera change is not necessary.
For more info, I’d suggest you look at the ‘making of’ tracks on any of the popular CGI DVD’s that have been released lately. On both the Monster’s Inc, and Toy Story DVD’s they have a good overview of how a scene goes from the layout->blocking->animation-final render stages.
I think the problem with most of us here at Blenderartists is that we do some major projects on our own. We’ll work for MONTHS on a movie, game, scene, whatever… when a team of somebody-else’s can get it done in a few days. I’m not saying that is the stereotypical normal, but having a team of qualified animators sure helps.
Then all the previsual department creating 2D moving images from the storyboard. I heard they used Blender to do previs of the NARNIA film. Heard about that at all?
Basically, can someone give me an overview of how CGI films are made?
mstram makes a really good point. I think people often think of the technical aspects in which CG films are different from ordinary films (live action or animated) and overlook the overwhelming ways in which they are the same. Sure you’ve got your modeling and CG animating and your compositing which is different from just shooting a live action film, but the skills that go into making those shots include lighting, blocking, directing, etc, which are all basically the same as in traditional film (except, thankfully, we can do crazy stuff like turn off shadows with the click of a button or restrict lights to a particular layer… luxuries not open to those poor saps using actual lights…) and once you’ve got your shots rendered and composited, you’re back in old fashioned movieland, where you have to string together a bunch of essentially unrelated moving images to make an engaging story emerge. Editing plays exactly the same role as in traditional movies, and if the director hasn’t filmed, animated, or otherwise created shots which can be edited together to serve the script (preferably a bang-up great script to start with, otherwise it’s all for nothing) then the editor will never be able to do a good job.
Basically, if you horse around with Blender for a few months and follow some tutorials, you’ll understand how modeling, animation, rendering, and compositing basically work. But this only begins to scratch the very surface of “how CG movies are made”.
I’ve looked at some major animation films lately, and you’d be amazed at the similar techniques they use. Bump maps… procederal textures, Ambient Occlusion… once you’ve learned the basics, it will help you no matter what you do.
