Benefit of using game engines for animated films?

ambelamba

 I know of at least one case:Boxcar Children. Looks like it was made with Source Engine or something and it looks...like a lot to be desired. I heard that it cost about $7 mil and wonder if most of the budget went for the voice talent. The movie was outsourced to some Korean studio (AFAIK) and I can't shake off the feeling that some of the intended budget was embezzled by someone.

 BTW, I was wondering if making some decent animated film with a game engine (in this case...obviously Unreal) is really feasible. I am not sure about the real benefit since I am no expert. Maybe they can cut down the rendering time and do a lot of scenes, if not all, with almost real-time approach. 

 My guess is that this might work well for TV animated shows. I don't know. It's my hunch. The whole point of using a game engine is to make the product look good with lower budget and shorter time, if possible. I am practically a layman but I am really curious how things will turn out if the whole production is based on Unreal and Nvidia VCA units. (is that even possible? I don't know.)

 

JordanN

I haven't made any movies, but I've used UE4 to make some personal cartoon projects of mine.

I think if you really want to, you can make an animated film in a game engine. I never felt limited in terms of kind of  content you can do in-game. Actually, it was quite the opposite. I love the freedom of being able to quickly move, edit or spawn any asset without having to wait on some huge ass render.

With a good enough budget, I don't see why it can't be done. 

Edit: Oh yeah, I also remember Epic themselves have shown off cinematics using just the engine and they're availiable for anyone to play around with.
https://www.youtube.com/watch?v=EO0k92iVMjE
https://www.youtube.com/watch?v=0zjPiGVSnfI

MagicSugar

For short films or series I think engines or machinima style of filmmaking is great, like for episodic series where heavy re-use of background or sets is not a big deal (e.g. Big Bang Theory).

For non-interactive feature length film I think it's still cost effective and more efficient to do it with the usual cg tools especially if all art assets are scratch built and not from pre-made packs.  Because for film there's less poly restriction, less texture baking that you have to do, no need to prep assets in the engine, digital mattes can take place over geometry, no need for engine concerns such as level load times, lod, re-targeting skeleton rigs, etc.  You don't have to render a scene with all elements intact, because you can render elements separately (foreground, background, vfx) and composite them in post. You could do "post" in game engines but is it as capable and feature rich as dedicated compositing softwares?

It's possible to produce a QUALITY feature length exclusively on an engine but I don't think it would be cheaper or faster than current methods and tools.

Kwramm

main advantage is faster rendering and quicker iterations. We did some research pre-UE4 becaue of that reason. Back then it didn't have the flexibility we wanted, so we shelved the plan. Another concern was that workflows, rigs, etc. would have to be changed and the film artists weren't too keen on that. But there's definitely an interest in this - especially TV cartoon show production would get a huge boost from it as quality isn't even close to feature quality, yet CPU bound rendering still takes a lot of time.