I optimistically hope that the new generation of gaming is more than simply a time for improved graphics. Some expectations are obvious, such as better lighting, smoother shadows and more of them. Likewise games shouldn’t simply look “more realistic” but should also look smoother and cleaner. Every game has room for graphical improvement.
I see this new generation as being an opportunity for so much more than just better visuals. This past generation, of the PlayStation 3, Xbox 360, and Wii, showed us that video games can to some extent reach the cinematic quality of a movie in terms of visuals and camera work. This new generation, however, should show us that video games, particularly game engines, can match the likes of a Pixar movie in terms of animation quality, physics, and overall impression.
By stating that I want games to be more similar to movies, I do not mean that I necessarily want more titles to remove control from the player or to implement overly linear level design in the name of “cinematic experiences.” Games are games and should stay that way, but there’s no denying that some games aim to be ‘interactive experiences’ instead of simply being games. In the past, however, one title aimed to be a movie before it became a game, and the day I learned this made me extremely happy.
Enslaved: Odyssey to the West, released in October 2010 by Ninja Theory, aimed to be a movie first. Yet it was sadly shot down by Hollywood in general.
“They have no idea of what real-time means,” he explains. “They just care about how good it looks.”
“Then, the fact that you say you can do this for cheap has no interest to studios whatsoever because they don’t want to do it for cheap. They want Pixar and the other higher end [studios] to do it for 100 million dollars or 60 million dollars.”
In short, the idea of using the Unreal Engine for a movie as was so foreign that the movie industry had no idea what to think.
I think a change is long overdue. We now have companies like NVIDIA who are a major part of the movie industry, using their technology to render frames in movies like Avatar (note that each frame still took two days to render).
Now amazing technology like ray-tracing, indirect lighting, “>real-time subsurface scattering, and even real-time hair, has made an appearance. Some amazing tech demos such as Philosophy, running the Luminous Engine, have been released to show at least some of this technology in action.
Video game engines are capable of making amazing computer-generated movies, and the demo above is confirmed to be capable of running in time on the PlayStation 4. That means what might be a $450 device (according to some UK pre-order sites) is capable of running a computer-generated movie perhaps comparable to or exceeding the visual eye candy of Fantasy VII: Advent Children at a frame rate of at least 30 frames per second.
Better Animation, Better Physics, Better Technology, Better Software
That’s all well and good: now that we have these great graphics and a higher poly count, everything can look sexy, but what’s holding all this back from taking over movie computer-generated imagery?
Some of the major factors are animation quality, physics, and the challenges with engines that are made primarily for making games, not for making movies. These factors need to improve greatly, and even though we’ve made some strides in some areas, these changes aren’t yet enough.
Animation is greatly lacking to start, in some cases being so poorly done as to ruin promotional images for games.
There’s no reason promotional images should have this problem. I originally saw this image in a blog some time ago that was illustrating this point. I unfortunately can’t find it, but the author demonstrated clearly that current game engines have significant problems with these factors, and game developers have to work really hard to make a character holding an object look true to life in current engines.
I’m sure you’ve seen this exact issue in TONS of games, where a character is holding an object, and it just snaps somewhere when they pick it up or put it down, or the object simply doesn’t seem to fit in the character’s hand. It’s odd, and it takes you out of the experience every time, even in cut-scenes where a delicate manual task is often hidden or just seems unnatural.
Then facial animation simply hasn’t reached its potential either: while a face can look good, even the best animation often comes out flat terms of expression. The problem, likewise, certainly isn’t that the animation is bad: some of these animators are REALLY good, but the final product is limited by currently available console technology. Current software can only handle so many key frames in a single animation. For this reason objects slip out of hands or snap, fingers don’t move like they need to, objects casually pass through each other, and facial animations can’t have the subtleties and detail they need.
What ends up happening is that fantastic animators with years and years of experience have to create a character animation that’s not even on par with what a student in a two-year class create. This is because the current software and hardware limitations simply don’t give them that creative freedom.
While the area of physics remains weak, we have seen improvements here. The cloth in Assassin’s Creed 3, for instance, was the first thing I noticed in the initial through. Something so small and seemingly insignificant as the cloth on a character’s coat made a great impression on me. It’s still not enough.
Cloth could be so much more: computer-generated characters in movies have base models with clothes on top to make a better impression. Clothing in games is static as it is: I greatly appreciate efforts such as ruffling said static clothing, but sometimes the effects make the clothes seem alive in the wrong way (“What is that thing consuming his neck!”). Clothing doesn’t have to behave identically to how it behaves in movies, but improvements in this area are things that need to be applied to characters and objects alike where necessary.
A great example of proper physics is the movie Adventures of Tintin, a movie that’s much like a European version of the Uncharted series of games, and to me this movie would have been perfect for being rendered in real time. The movie contains fantastic animation, wonderful cloth simulation, and TONS of particle effects. I can’t see previous-generation consoles such as the Xbox 360 as being capable of managing these details in real time, but the new generation of consoles will certainly be capable of doing so–and they must do so.
Major changes are necessary in the technology being used. The PlayStation 4 and next Xbox seem to be the driving force for this new technology. Thanks to their adaption of a more PC-like architecture (x86), we can see this tech spread around. At this point, however, technological improvements seem merely visual.
One demo, however, seemingly show all these details in place and comes the closest to Tintin that I’ve seen. Physics and particles appear when necessary, and every character has cloth simulation: the best part, however, is the animation. The elder in the demo actually has believable wrinkles and complete facial animation. There are subtleties here that you won’t find in the previous generation of games. The demo is a visual masterpiece, to be sure, but the technology is a perfect example of what animation needs to be in games.
On top of this, the tech demo shows that this technology is something easy to use: it seems fully capable of rendering movies as well as games. To be honest I wouldn’t be surprised if Square Enix’s efforts in movies went to their own game engines in this same way. They’ve proven capable of producing a product fit for the big screen. A more powerful computer could add even more resources for them to take advantage of.
Square Enix isn’t the only one: Unreal Engine 4 has certainly made strides in this kind presentation, and the engine powering the Second Son trailer proved that too.
I believe the new generation of gaming will be a time for movies as well as games. There’s a lot to be gained here, and video game companies have already proven they have the potential and will to make a movie. I don’t expect the movie industry to take advantage of this yet, but a movie made with a game engine like this is low-budget and easily accessible. It could even one day be the standard for short computer-generated films and television shows.
This effort into movies comes back to games as well: if video games companies begin working in areas other than just games, we could see a surge in quality and content. We could see more effort into improving technology, and anyone who worked in the game industry could also take part in the movie industry too.
This new generation has the potential, in my opinion, to change the industry. What do you think?