Escapist Podcast: Is Photorealism Needed for Gaming to Advance?

[The Escapist Staff]

Is Photorealism Needed for Gaming to Advance?

This week, we discuss 2K Games' comments about gaming needing photorealism to mature. We also discuss game localization and the announced third Hobbit film.

Watch Video

 

[MagmaMan]

No, it is not, in the slightest! I want scale to increase! The current visuals we have are fine.

 

[DrRockor]

art style is so much more important than graphically fidelity. Getting closer to true photorealism is awesome but there are other styles that don't need the same level of graphical power and give the same emotional experience.

 

[Jimbo1212]

Yes it does if only to make the game more immersive and appealing to a larger market, but the key is that photorealism = more power from the system, ergo the game can have more mechanics, physics, be larger etc.

No, it is not, in the slightest! I want scale to increase! The current visuals we have are fine.

That's just rubbish! Many games already suffer from a lack of graphical fidelity, be it mistaking a guy 100m away for being part of a pixelated blur, or just good art design going to waste.

art style is so much more important than graphically fidelity. Getting closer to true photorealism is awesome but there are other styles that don't need the same level of graphical power and give the same emotional experience.

Such as?
Just remember that immersion =/= escapism thus photorealism and art styles will not give the same results.

 

[ReiverCorrupter]

Yes it is. These guys are way too focused on emotional involvement. Let's put the whole "what do games really need to be artistic and meaningful" thing and just focus on the fun interactive side of games for a second.

Photorealism is necessary simply in terms of what it will mean for the division of time and resources in game development.

You might hate the AAA FPS industry, and that's fine. But like it or not that's the genre that drives the industry. The best selling games are FPS, and they are what take up the largest amount of the industry's resources. I'm not saying that it's right, I'm just being pragmatic. The fact of the matter is that the entire industry benefits from the technological innovations produced to make the AAA FPS titles, e.g., the Unreal engines. It also drives the trends.

Developers spend these resources on photorealism because it sells. The reason it sells is because graphics are important for the gameplay mechanics of FPS games, much more so than other genres. Just think about realistic foliage and camouflage affects gameplay in an FPS. The graphics are usually a selling point for FPS games because it shows that the game is polished. It's the most obvious way to compare games.

What this means is that the majority of the industry's resources is put into making better graphics engines and not, say, better AI or an economic simulator. Take Skyrim, for instance. It could have really benefited from better AI, or maybe an economic system that allowed your player to build a business empire, or hell, better combat mechanics. Instead they spent most of their resources on making a giant beautiful open world because they knew that would be the selling point for a lot of their audience. Once they no longer have to spend most of their time and resources on painting plants and mountains the gameplay mechanics will start getting a lot more innovative.

It just comes down to the development process. Designers have finite time and resources, the more they spend on graphics, the less they spend on gameplay. Once game developers get design kits that allow them to easily build a photorealistic world they'll be able to spend more of their time on game mechanics. Games will have to compete in terms of gameplay mechanics, and this will cause a new wave of innovation.

To put it simply: when designers no longer have to worry about graphics, they'll have nothing left to do but put all of their effort into making the game fun.

You can complain that that's what they should be doing already, but that's just unrealistic. The fact of the matter is that people would have complained endlessly if Skyrim didn't look a lot better than Oblivion. Many of them would no doubt take it to mean that the game isn't very good. Are they ignorant basterds? Yes. But they drive the industry.

 

[RubyT]

Of course it's not "necessary". Otherwise the video game market could have never emerged. Somehow Mario and Tetris were fun enough.

True advancements can only come from better AI. Smarter NPCs and smarter story-telling (i.e. more flexibility and freedom).

But GPUs have just increased their power so much over the last 15 years and it's so comparatively easy to use the parallel processing power of GPUs for improved graphics fidelity, that I understand why the industry is focussing on it.

Nobody knows how to properly write an AI.

It has to be said though, with all the money and effort they put into Crysis, for example, it's shocking how they cared NADA about the writing. The "screenplay" of that game seemed to be written by two 8th graders during lunch break. How expensive can a somewhat experienced writer from the video game industry be? I don't even think they'd make up 1% of a AAA game's budget. Is it really the place you'd want to cheap out.

Why yes, yes it is, what with the massive success of Farmville, Angry Birds, MineCraft, World of WarCraft and all those shooters who only live by their multiplayer-modes, it's not hard to understand why upper management thinks players give a crap about story.

 

[PureIrony]

I know its kind of pedantic to stop listening to a podcast and correct the speakers, but the children's book you were all thinking of is The Twits, which was about an elderly couple who constantly played awful tricks on each other(putting worms in the other's spaghetti, trapping birds to put in pies, etc). They were killed by vengeful crows who superglued them to the ceiling.

So...yeah.

 

[ReiverCorrupter]

Yes it does if only to make the game more immersive and appealing to a larger market,

Huh, that's another point that I hadn't even considered. You're right. I think part of the reason why a lot of people see games as children's toys probably has to do with the lack of photorealism. They perceive them as being cartoonish. Sure, we can call those people ignorant but the fact of the matter is that the more people who buy games the more the industry thrives and the more we all benefit.

but the key is that photorealism = more power from the system, ergo the game can have more mechanics, physics, be larger etc.

Well... photorealism is a result of getting more power from the system, not the cause. We improve the processing power of our hardware regardless of what we use it on. Thus far we've been spending greater and greater amounts of processing power on graphics engines. So I'd actually say that the pursuit of photorealism has hindered the power of video games by taking up too much their engines.

But you're right in that once we've reached photorealism each proceeding advance in computing power will have to be dedicated to something else aside from graphics. This will force developers to develop more and more innovative gameplay mechanics.

When we reach photorealism choosing video games will be like choosing between supermodels; they're all gorgeous so you might as well date the one who's the most fun.

(Hopefully no one will be overly offended by my sexist analogy ;P)

 

[ReiverCorrupter]

Of course it's not "necessary". Otherwise the video game market could have never emerged. Somehow Mario and Tetris were fun enough.

True advancements can only come from better AI. Smarter NPCs and smarter story-telling (i.e. more flexibility and freedom).

But GPUs have just increased their power so much over the last 15 years and it's so comparatively easy to use the parallel processing power of GPUs for improved graphics fidelity, that I understand why the industry is focussing on it.

Nobody knows how to properly write an AI.

That's because AI is incredibly difficult. In fact, there's still a lot of debate over what AI would really be and whether it's even possible by simply writing enough code. A lot of people think (particularly in robotics) that AI can only be possible by setting up hardware that is capable of rewiring itself and learning from its environment, much like an organic brain, and that trying to get true AI by writing software alone is futile.

However, the reason the industry is so focused on graphics isn't just because it's easier than writing AI. I think you underestimate the importance of graphics to the AAA FPS market, which is kind of the driving force behind the industry. Sadly, people who buy CoD and other such shooters do tend to focus on graphics such that it has become a major selling point. The people running EA and Activision think graphics will be the deciding factor that will get people to buy their flagship shooter instead of their competitor's (stupid though it may be.)

Another important thing to remember is that pretty much every game can benefit from improved graphics, whereas not every game utilizes AI, so it's probably a better overall investment. Note that you can make almost any type of game using the Unreal engine, but your AI is almost always going to have to be written to specifically fit the game your making. The AI from Deus Ex is completely different from the Radiant AI in Skyrim.

You're right in that the true advancements that will make video games better will come from better AI. The problem is that people won't focus the proper attention on writing better AI until this stupid graphical arms-race reaches its photorealistic conclusion.

 

[Eric the Orange]

I probably would have been afraid of your dog up until my teen years when I grew out of it. But for a time I was petrified of all dogs.

Also I find the translations of the persona games interesting because they are so immersed in Japanese culture. Japanese holidays, Relationships at Japanese schools (like upper class men and lower class men). Because it's meant to take place in modern day Japan. So it's tricky to try and bring across these things across that are very foreign to a western view point.

 

[RubyT]

Another important thing to remember is that pretty much every game can benefit from improved graphics, whereas not every game utilizes AI, so it's probably a better overall investment. Note that you can make almost any type of game using the Unreal engine, but your AI is almost always going to have to be written to specifically fit the game your making.

Very true.

There's another big problem with AI: it's difficult to test. The more you limit the freedom the player has, the better you can control how the game handles his interactions. If you incorporate a good AI and give the player appropriate freedom, it's hell to ensure your game mechanics hold up.

And if you'd write a relatively strong AI, you'd also face the problem that it will end up to be non-deterministic, i.e. you won't be able to test at all if it will always react appropriately.

But I'm not even talking about this kind of AI. There's a lot that could be done today, with simple scripting, but it's a lot of work and currently, the game design department isn't receiving the same budget as the tech department.

You're right, AI will not get the money and brains it deserves until graphics have a reached a point at which the John Carmackss of the industry decide it's no longer worth their time and join the Peter Molyneuxs.

 

[ReiverCorrupter]

Another important thing to remember is that pretty much every game can benefit from improved graphics, whereas not every game utilizes AI, so it's probably a better overall investment. Note that you can make almost any type of game using the Unreal engine, but your AI is almost always going to have to be written to specifically fit the game your making.

Very true.

There's another big problem with AI: it's difficult to test. The more you limit the freedom the player has, the better you can control how the game handles his interactions. If you incorporate a good AI and give the player appropriate freedom, it's hell to ensure your game mechanics hold up.

And if you'd write a relatively strong AI, you'd also face the problem that it will end up to be non-deterministic, i.e. you won't be able to test at all if it will always react appropriately.

But I'm not even talking about this kind of AI. There's a lot that could be done today, with simple scripting, but it's a lot of work and currently, the game design department isn't receiving the same budget as the tech department.

You're right, AI will not get the money and brains it deserves until graphics have a reached a point at which the John Carmackss of the industry decide it's no longer worth their time and join the Peter Molyneuxs.

Agreed. (Although hopefully they'll join someone a bit more sane than Molyneux.)

But I wouldn't just blame impoverished AI for the increased linearity of games. The constraints placed on them by voice acting play a huge role as well. Just look at the depth of different options and conversations in a text based game like Morrowind and compare it to Oblivion or Skyrim.

Games won't even come close to passing a Turing test until they can generate original dialogue. And that not only requires an amazing AI that can respond intelligently to what the player says, but also some way to generate human sounding voices that carry the proper emotions and intonations. I'm afraid that both of those things are way further away than photorealistic graphics.

 

[Susan Arendt]

I know its kind of pedantic to stop listening to a podcast and correct the speakers, but the children's book you were all thinking of is The Twits, which was about an elderly couple who constantly played awful tricks on each other(putting worms in the other's spaghetti, trapping birds to put in pies, etc). They were killed by vengeful crows who superglued them to the ceiling.

So...yeah.

Hmmmm....maybe, but I don't think I've heard of that.

 

[Rakor]

Atlas, Disgaea, hells yeah. Nuff Said.

Nin nin nin.

 

[PureIrony]

OK, now they I've actually listened in, let me make a counterargument.

I think we need new mechanics, not better graphics. Heavy Rain should be more engaging, should be playable in a way outside of just multiple QTEs and pressing the x button at things. I'm convinced there's a way to make games like that, if not fun, at least less of a chore.

 

[Squilookle]

No, photorealism isn't necessary and never was. Too much focus is on graphics already, when what we really need to be catching up on is A.I, and storytelling for the games that choose to employ it.

 

[gardian06]

oh my gosh. what is going on here? Graphics get money because developers think that if they make the game pretty for the screen shots (and the rest of the game) then players will not notice that the number of mechanics, and features is lacking (yes they are 2 different things any 1st year game design student, or novice game tester can tell you that).

and the argument that the game engine determines the amount of graphical quality. I have been working with game engines like Unity, Unreal, and so on for a few years now, and would like to ask the question "where is this button that makes my graphics better I keep looking for it and can't find it anywhere?" oh wait it doesn't exist the quality of graphics is based on the system that is running it (partly), and on the artist that is rendering, modeling, skinning, rigging, animating, placing the object into the game.

the reason art (interchangeable with graphics) gets so much money in a game is 2 fold. 1st its because developers are given tech specs for a system (this counts for PC to), and get all gitty like a kid who just got a new toy, and wants to see everything it can do. this leads to such a focus on art (because that's all they think all the processing power is good for) that whenever the subject of new features comes up that is just pushed to the back burner so it looks cool.

2nd because artists are expensive: that list up a little bit that's not an or list its the step process, and that process gets expensive quick. I will not give a break down here, but Master Chief probably costs a good million (severely low ball estimate for mediocre artists) every time he is redone, and that's not talking about any of the other characters that actually have faces which shoots the price up drastically. So the next time you hear about a games budget think to yourself that 70-80% of that went to various artists, and the rest went to the development of the actual game.

When it comes down to it games don't need realism. because Jim was right when he said the best looking game was Viva Pinata of this gen because for all that money that goes into "realism" while graphics continues to improve (after deep breath) those games will look like shit in less, and less time.

when people say that "the graphics arms race will end" this will only be true when companies like nVidia, and ATI say the words "we have enough money", and the big 3 say the same because tech demos of processor speed get developers thinking, but tech demos of poly count those make developers wet themselves with glee.

the big reason graphics gets the big focus in game development is because it take the least effort on the part of the company, and when it comes down to it its the most easily recognizable.

what I always find funny are those people who complain that "games are to short for the money" here's a fun one go, and look up the amount of space on a disk it takes for just one frame of "hi-def" video, and then think for complex (number of moving points on the object) animations that can actually be more. so if a game has alot of hi-detail-complex animations, and even an hour of video (opening, cut-scenes, ending, loading cinematics) you can quickly figure out why your game is only 8 hours on a DVD, or even a BlueRay

 

[Dastardly]

Is Photorealism Needed for Gaming to Advance?

This week, we discuss 2K Games' comments about gaming needing photorealism to mature. We also discuss game localization and the announced third Hobbit film.

Watch Video

Interestingly enough, current research is showing that people more quickly recognize caricatures of familiar people than actual photographs. The hypothesis is because these exaggerated, "stylized" images might be actually more like the way the brain actually stores that person's image.

 

[GamerLuck]

So... Can we get a picture of susans dog? we send er cute things all the time, she should post her cute dog

 

[Koshok]

I'd like to response to the localization discussion. I'm sure I'm in the minority when I say this, but I would prefer my media to not be localized at all. When I used to watch a lot of anime, I really appreciated when the translater would write a sentence that may not be fully translated, then give a note explaining it.
For example, if a character makes a pun, the joke is very dependent on the language. If the pun gets translated, it either loses all meaning or the translation has some absurdly contrived sentence to reflect the fact that the character made a pun. I would rather have a more accurate translation followed by an explanation.

Though I'm not sure if this would work for games as much. I really wouldn't want to pause what I'm doing in a game to brush up on a culture's eccentricities.