Slashdot Mirror
What if Game Graphics Never Aged?
An anonymous reader writes "If you've heard of Procedural Synthesis, you already think it's amazing. It's been used to create some extraordinary visuals in tiny packages, like .kkrieger, which is less than 96 Kilobytes big but still has graphics that look like like a modern PC title. Beyond that, there's even more that Procedural Synthesis might be able to do; what if your old video games never aged, never looked out-of-date? Imagine putting Halo 2 into your Xbox 360 only to have it automatically upgraded to look like Halo 3 in graphical quality. This article examines the unexpected way that Procedural Synthesis might impact gaming in the generation after the Xbox 360, PS3, and Nintendo Wii."
There's all this hype on graphics and technology, but the heart of any game is still (and always will be) gameplay. Sure the games of old look "crappy", but in many cases they provided a great gaming experience. I for one hope that we just get to the point where graphics are real-life quality and we can focus on gameplay. Just my $.02
http://religousfreaks.com/It would be nice if they incorporated some of this technology into the Wii. The old games are great but they would be even better with the graphics turned up to today's standards. I've played with a few emulators that added Anti-aliasing to old SNES games and such. The games looked a lot better. I recently bought a new computer, and hooked up my old copy of Descent 3. It still looks amazing. This is because I was able to turn up all the effects to the max. Whereas before, I was stuck with everything at half. I'm sure the same thing still holds for video games. Most people can't play new games at full res and full effects. However, in 5 years when they buy a new computer, it will be able to pull it off easily.
Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
.kkrieger is certainly a feat of software engineering (pretty much anything .theprodukkt puts out is) but procedural synthesis can only go so far. When you get to elements of the game that should be static (such as specific characters) then a static model would probably be more efficient than an algorithm to generate the same.
Of course, I could be (and probably am) wrong.
A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
That's an interesting thought. The article makes it out to be a bit like a magical cure, but some aspects of it sound good to me. You can often improve the "wow" factor by tossing in "more" of something. Denser foliage; more of the tiny rocks that make up the detail; and so forth. Procedural generation would mean that these wouldn't have to be placed by hand, so this could make it easier to scale the visuals with system power. Similarly, particle sprays are often done procedurally, so being able to tweak those "up" to create more complex fireworks for mysterious future hardware could also work.
:)
Some games are still played for years after they've fallen behind the curve on graphics; this might mitigate the future ugliness, adding longevity to a popular title. Keeping gamers interested in (and talking about) your game makes sense, whether you'll be producing different titles in the future or will be focusing on sequels.
Ultimately, though, my hope is that algorithmic content generation will bring game development costs down for indies. Maybe I'm dreaming.
_______________________
Indie Superstar - A video webcast for gamers who play indie games
Dejobaan Games - Indie games for people who watch video webcasts
We're indie. We're working on our 14th game.
There's more probability of that if the graphics automatically upgrade on new hardware. It would make the graphics less of a selling point.
Tharkban (It is a signature after all)
Exult was a good example of "procedural" "growth" of a game.
Ultima VII was a 2D RPG. Yet, all objects in the game world have height. One guy at Exult hacked up a version of Exult that runs Ultima VII in 3D mode - basically, mapping all 2D tiles around cubes as described by their dimensions and height data.
The results were quite interesting (buildings looked kind of good, creatures and many plants and natural formations not so good, so they are being replaced by 3D models).
But it is a good example and exercise in extracting more detail from the game than the original developers intended or envisioned.
If those old NES games looked like modern games, I wouldn't play video games at all! What we need is the opposite, something that makes current games not look crappy and 3D.
This would be AMAZING, though, as an open-source project. Get an amazing, constantly-updating engine down, and let people release all the content they want for it. It would be like Doom WADs on steroids.
Firstly, there is the question to how much effort a company would want to put into making a version of their game that gets better with age. Using current models of "create a game, sell a couple hundred thousand copies, then make another game" it doesn't really make sense. The key is that the graphics can improve as hardware improves, and the only sorts of games that really come close to fitting that sort of lifecycle today are the MMOGs. Like I wish Ultima Online had graphics that had improved over time. The game is almost 10 years old and is largely unchanged. Other games (like the soon to be serialized Half-Life 2 and SIN series) might also benefit from improvements, though HL2 is already incorporating improved graphics with each new episode (according to the developers commentary). Secondly, the procedural systhesis method is much more compute intensive. They use as a prime example the forest scenery in Oblivion. As we all know, Oblivion is a performance-killing game on the PC, and the Great Forest part of it is the slowest part by far. So if you go too far with procedureal synthesis today, your game can turn out to be a real pig. So there's a definite balance that you have to strike between performance today and upgraded visuals tomorrow.
At the very least, you can use this to make distros easier. Think of all the different Source games from Valve that are available for download. I think it'd be pretty cool to download HalfLife 3 in a couple minutes on a dial-up. You could do that if you rendered the content, and could even set it up so you only had to do it once on install, and after that it would be no different than if the content was rendered by Valve or by my PC. So the load time is really just install time, which you probably already put up with now. Just a thought... Cheers.
I would say that GTA actually has way more "plot" that HL2. HL2's plot consists of "make your way to point y to meet person x, killing lots of zombies along the way". It's fun stuff, but it's not exactly an involved plot. Although I'm only a small bit of the way thru GTA, it actually seems to build a bit of a plot line with some questions about the various characters (e.g. Smoke seems awfully chummy with the cops).
Some game companies (including my employer) seem to think that their games are poor quality (oops...better not reveal my employer's name) because the storytelling isn't good enough. These people look to the movie business and see that many big effects movies suck because they have a weak story and assume that the same criticism carries over. It doesn't. Games and movies have a whole lot of different ways in which they can suck that don't relate to each other.
The game business seems to look to the movie business as a kind of more respectable big brother. So many game developers have now got it into their heads that they must try to develop things like movies. And hence they feel pressured into developing a story even though they may end up wasting resources that might better have been used for gameplay.
A nice example of the latter is the old adventure game business. Because these game developers felt that somehow what they were doing was lowbrow they renamed the genre to "interactive fiction" denying their games heritage.
Make games and be proud to make them, whether they have great graphics, great stories or great gameplay. Don't feel that somehow you have to compete with other art forms like literature and cinema on their own turf.
Doesn't it make you feel good to know that our freedoms are protected by politicans, lawyers and journalists.
Definately agree. Choose which segment you want to be heavy in and go for that audience. That doesn't mean if you're making story games, ditch gameplay and graphics, but rather find a good balance, and when necessary, go heavy where your demographic requires.
Games shouldn't compare themselves to the movie industry as they are two different forms of entertainment, I agree with you there. It may be possible to take some cues, but anything more doesn't work (lighting, camera angles that are absolutely beautiful on film tend to come off slow and boring in the interaction-based game medium). Nevertheless, if you're making a game with a good story, I don't feel that story should be compromised for "gameplay reasons" and in most cases it doesn't have to be. It's very possible to have both once a studio gets out of the we-must-be-like-hollywood mindset.
That is, if you're in the business of making plot-based games. If your market is Tetris 5,000 or Breakout: The Latest Iteration, then for the love of all that is holy keep that plot away from it, because it just comes off cheesy.
Procedurally generating content is computationally expensive, and not always easily done in parallel. To procedurally generate an entire level worth of content would take a lot more time than most people would wait for a level to load, for even a relatively simple and small level it wouldn't be surprising for it to take a few hours- even with heavily optimzied code.
Things like texture synthesis can be done on the fly for simple cases, and procedurally generated models are being used today to some extent, but generating everything procedurally won't be realistic for quite a while.
The other big thing is that procedurally generated content is rarely perfect. There are some things that can be done really well, but most things require a lot of processing to minimize the weirdness that can crop up from time to time.
What it seems more likely to me that would be done is, instead of procedurally generating content on the console when the game loads, game makers will start to look more at procedurally generating content during the development process. This can help bring down the art requirements on large projects, and allow greater detail without requiring things to be run in real time.
Famous Last Words: "hmm...wikipedia says it's edible"
I'm not really sure that there's a problem to be fixed. The modern console RPG is a totally ligitimate form of entertainment, IMO, that some of us enjoy even more than tabletop games. Though, I will agree that they are far removed from their origins, they've become something else, entirely. And yes, they probably should loose the "RPG" title, but I have yet to hear another title that didn't sound either incredibly stupid, or incredibly smug (ie: "interactive story book").
I, personally, am perfectly happy playing through a pre-defined story. This is the way that narrative entertainment has been from the beginning of time, and it's been perfected throughout the ages. I'm not saying that there aren't other ways, but to say that it's a "problem" or in someway inferior, is just one ignorant opinion.
But I can understand that many people are looking to the table-top RPG, and are interested in trying to figure out a way of replicating, at least partially, that type of narrative experience in a video game. And I am currious as well. But until we can create AI that is nimble enough to create engaging narrative on the fly, we only have 3 options: create a pre-defined storyline in which the characters can play through, create a branching storyline in which the characters actions define the particular path, or completely give up on story altogether, and concentrate fully on the hack & slash model, which I don't believe to be any closer to the table-top RPG experience than the pre-defined story model.
Now, I highly disagree with you on the notion that tabletop RPGs are more about hack & slash than plot. If all you've played is D&D with GMs who're only interested in fighting, then I can understand how you might think that. I've played whole adventures which consist mostly of going around and talking to NPCs, or figuring out the history behind situations and creating a new dramatic plot as the game progresses. I had one GM who ran a 2nd edition campaign which focused, almost entirely, around the international politics of the region we were playing, it was one of the most fun and exciting tabletop experiences I've ever had. I've played Dark Conspiricy campaigns that revolved around ELABORATE narrative and interactive story-telling. A friend of mine created a tabletop called Aeon, and the beginning adventure involved no fighting at all, mostly a lot of detective work. Hell, there are thousands of people who play various LARPs, most of which contain only the very rarest instances of fighting. I played Vampire for a whole year, and there was only ONE mass battle, and everyone agreed that it was really pointless. No, the reason why the manuals of tabletops are filled only with stats is because storytelling doesn't need instruction. GOOD storytelling takes years and years of practice, thought, and study, but one book isn't going to be able to help much. Most GMs would rather spend their time on creating narrative then creating a system; so the game company supplies the system, and the GM supplies the narrative. In that case, one could argue that tabletop games revolve MORE around narrative then systematic action... but every player and GM has their preference on balancing these two interests.
Multiplayer Gaming (defined): Sitting around, discussing single-player games with my friends, at the bar.