The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.
Maybe I'm spoiled or jaded, or maybe I am just searching for something different these days but for all the hype and leaked specs I'm still not excited about the next generation of consoles. I don't know why, but I feel like I should be more excited by now and still, nothing.
I recently read an article on Wired that unveiled Unreal Engine 4 to the public and my immediate response was, "meh". I know right? Kind of shocking that 10 years have passed in this current generation of gaming and somehow the next big thing looks about as good as the thing we already have. I read the article, and there were some definite sales points; all dynamic lighting, some bounce light calculations, and likely a number of other image-based approximations. The article went on, plucking quotes like the push to render Avatar in real time. That goal is fine and all but the screenshots did not really paint any of the things that were being discussed in the interview. To me, the screenshots looked like UE3 with particles...
Epic Games has a tough sell this next generation. We aren't seeing the generational leaps like 2D to 3D, or fixed-function to programmable GPU's. The next "leap" for Gen4 is... dynamic tessellation, and more GPU cores for more complex shaders. Not exactly the bullet-points you write home about. For a long time the next generational leap was predicted to be real time ray-tracing. Many of the image-based hacks and post-process distortion effects we do would have disappeared in favor of an unbiased tracing technique. If graphics is the foot that we want to start with in Gen4, perhaps seeing pixel perfect fuzzy shadows and volumetric illumination across a heavily vegetated landscape would have really made a strong argument.
I recently finished Deus Ex: Human Revolution, which is an amazing game by the way. Aside from some really bad stereotyped civilian characters and some weird bug that forced you to read emails twice to activate security codes, the game was a great experience. This game was also a strong case for a Gen4 console. There were inexplicable breaks in the world that would force a loading screen while other areas would briefly hitch, telling me that they were likely streaming some parts of the world in and out. I assume that the developers were hitting system memory limits, or perhaps too much CPU was being consumed by AI that was clear on the other side of the map. Whatever the reasons, increasing CPU and memory would have done a lot to improve the experience of this game. Graphically, the game was very smartly designed and every room was well lit and blooming like J. J. Abrams himself blessed the development team. It's clear that better definition could have helped solidify the characters into the world but I get the feeling that the quality of the characters was a budget constraint more than a technical one, given the large number of characters and variants in the game.
Deus Ex is a strong case for a Gen4 console, but not an easy sell. Assuming all of the adjustments I mentioned were a product of limited hardware resources, I don't know if those changes would have turned Deus Ex into a system seller, a reason to buy a $600 console when that $200 console does the job pretty good. Can graphics really sell a console anymore? Or is it really aesthetics that will sell the next generation?
The funny thing about aesthetics is that we have the capability now. As I mentioned, Deus Ex was a very good looking game. We also have games like Journey, Limbo, or even Braid; games that seemed to look beyond the graphics and focused on the aesthetic. It is difficult to say if these games would have been any better with more horsepower, or if that power would have simply been used to make their development schedules easier by abusing the abundance of computing power through rapidly developed, poorly optimized systems.
A resurgence of 2D games in all genres leaves me asking the question, "is real-time Avatar what we really want?" It takes 200-300+ people to put out a Call of Duty or a Grand Theft Auto, ballooning budgets over that $100M mark. If it continues to take roughly X hours to get Y done in game development then are we going to grow teams to 500 people and stretch development to 4 year cycles? Can a $350M video game even make their money back?
The reason I raise those questions is because it comes full circle to the Gen4 dilemma. If graphics is what sells consoles, and Gen4 is more of a Gen3.5, and there is a fixed cost in man hours to develop features, research, and innovate; what is the strong sell for the next generation?? More horsepower is good, but it may end up going to the tough sellers like better sound mixing, more cycles for AI, more iterations for physics simulations, ancillary animation systems, real-time subsystems for simulating world events. None of these things are features that gamers can show their friends in a screenshot. And even if they could show an amazing screenshot, what are the chances that the viewer could possibly understand that what they are looking at is "better" when we are still using the same fundamental techniques for creating and rendering content? As it has been for some time now, the power really seems to be in the hands of the artists with the skills to exploit the basics; solid modeling with clean and colorful textures.
Gen4 will eventually come and overtake Gen3, but I get the feeling that artistically smart games are likely to overtake the market, not graphically superior games. As I said before, Epic Games may have a tough time selling UE4, but only because some developers may choose to use "aged" engines that continue to exploit a more unique visual style and tactile feel than a licensed solution.
All around, Gen4 is going to rock some large boats...