Why Haven't 3D Graphics Surpassed 2D Game Art?

Thanks to GameSpot for its 'GameSpotting' article discussing the longtime game player's "soft spot" for 2D games, and why, in the author's view, "3D polygonal graphics still haven't entirely surpassed 2D game art." He explains: "In a way... I think the cinematic power of gaming almost took a step back with the transition from 2D to 3D. 2D game characters are displayed precisely how the artist chooses to display them to you. There is no extraneous frame of animation to be found. 3D game characters, meanwhile, are yours to control, so you may rotate them and view them from whichever unflattering angle you like." It's also argued: "2D games handle collision detection (or the interaction between two characters or objects) better than 3D games do... [and] I think 2D game characters still have the capacity to display more-lifelike emotions than 3D game characters do."

Comments

It's the level of abstraction

[foidulus]

kind of like the difference between animation and live action movies/television. There are certain things you can do in animation by not making the characters concrete. A live action family guy would probably be disturbing, much as the live-action tick is. The charm in those shows was that you were dealing with somewhat abstract beings. The same thing goes for games, for certain games, I just want an abstraction of what is going on, ie 2d, I don't want something that is nitty gritty realism. For others(doom 3!) that works out really well. I think it largely depends on the game itself and what the authors are going for.

I'm a bit resentful of the 3-D platform.

[hal2814]

I was a huge fan of the 2-D platform games. While I do think that there were planty of bad 2-D platformers, but I really think that they had a much larger percentage of good games than most genres. When the 3-D platform games came out, it's like all game companies just abandoned the 2-D platform. The gaming companies saw 3-D platforms as an upgrade instead of seeing 3-D platforms for what they are, a different gaming style. Soon Mario, Zelda, and Sonic all left the 2-D arena (except for a couple of subpar 2-D games like Yoshi's Island and some Sonic compilations of older games). Then everyone else followed. 3-D platform games should never have been seen as a REPLACEMENT for 2-D platform games, and that makes me resent most 2-D to 3-D conversions somewhat (ESPECIALLY the 3-D bastardization of Bomberman on N64). That doesn't mean that there aren't 3-D platform games that I like (ex. Sonic Adventure series). It just means that I've got this unintentional bias towards the 3-D platform games that I can't always seem to get past.

Re:I'd have to disagree about collision detection

[Anonymous Coward]

It's not so much a problem for moving-around collisions any more. Sometimes you run into situations where you can't fit through something you should be able to, because it's got diagonal edges (like a crack in a wall) and your player collision model doesn't (it's square). And very occasionally the reverse will happen. In FPSes especially, game designers "cheat" and make crouching bounding boxes smaller, even smaller than the player's visible model, so they can avoid being trapped on the scenery. This is why you have things like guns which poke through walls and even boosting through floors/ceilings.

Anyway, all that's more a nuisance than anything else. It's not a problem in the way I think the OP means, which is getting stuck on stuff, cause that seldom or never happens any more. The real problem are hitboxes: the region where you can be shot and have it count as a hit. Again, these are usually a couple simplified boxes which correlate roughly to a player's appearance. The problem is that they suck horribly. I haven't played a single game with even remotely accurate hitboxes, and the main reason is that boxes don't curve. Consider a human body in profile. The head is a little narrower than the body, right? So if you think in terms of boxes, the head hitbox will be thinner than the body hitbox. But because they're square, if you're crouching, right in front of someone who's standing, and you aim up at his head, you're going to hit his chest. Even though you have a clear line of sight to his head. This is because the top corner of the chest hitbox is sticking out in the way, and your bullet hits that first. There are all sorts of annoying angle crap issues like that which bug the hell out of anyone who plays CS (the gold standard of horrible FPS hitboxes).

The good news is that D3 claims to be using to-the-pixel hitboxes, that is, not hitboxes at all. This is probably also why they say you need 100mbit/s to get more than 4 players in a multiplayer game. But still, it will be very nice if true, and if it extends to the model collision detection (for stuff other than bullets) too, it opens up all sorts of modding possibilities for crawling around.

Why this article?

[Sigma 7]

I'm not sure what the author is talking about, and as a result, I'm not sure why this article was selected. For example:

, I was reminded that 3D polygonal graphics still haven't entirely surpassed 2D game art.

In this summary statement, the author himself states that he is comparing basic graphics to art. Nobody cares that bout this - what needs to be done is either a comparision between 2d graphics and 3D graphics, *OR* a comparison between 2D Art and 3D Art.

My problem with 3D graphics in games--and it's always been my problem with 3D graphics in games--is that they're unedited. You can often view the action from any angle, and frames of animation are typically never skipped. In a way, then, I think the cinematic power of gaming almost took a step back with the transition from 2D to 3D.

The only advantage of 2D movies is the fact that you can draw fancy art to as high as a detail as you want. 3D sequences, while not looking as fancy, do not require as much space as their 2D counterparts (by reusing models, textures and so on), and can be consistantly modified without having to redo many frames of work. Also, I am finding that modern games have cinametics comparable to how it should look like - it's a big jump from Dark Forces (an old Dos game that used simple cinamatics) or Jedi Ourcast (3D cinamatics don't look ultra-fancy, but get the job done.)

Not only that, but there are ways to convert 3D-graphics into pre-rendered 2D movies without problem. From there, it's quite easy to do the "editing" that the author seems to want. Not that it matters, since I have very rarely seen an issue with 3D graphics in the games I've played. The closest thing would be those "classy" screenshots posted on PlantUnreal, and those could be pulled off in a 2D game with the same complexity.

Besides, the author ignores the "rotating-corpse" issue that was visible in Doom where you could only see one side of the body after it was killed.

I still think 2D games handle collision detection (or the interaction between two characters or objects) better than 3D games do, on average. And having good collision detection is one of the most fundamentally important aspects of just about any game. Likewise, I think 2D game characters still have the capacity to display more-lifelike emotions than 3D game characters do.

This is easily countered by using Wing Commander 1 compared to X-Wing. While X-Wing might not have looked fancy, you could easily tell when you were about commit suicide by ramming a Star Destroyer. In Wing Commander 1, the collision box was independant of the sprite, and you could thus accidently bump into a Ralari without knowing it (not only that, but the collision box was based around a static box rather than the visible model/sprite.)

Now the other problem with collision detection in 2D games - in the games where collision means death, you either have a per-pixel collision detection, or bounding box collision detection. In the former, you die as soon as one pixel nicks whatever you are supposed to avoid. In the latter, you can't tell if that tight squeeze is fatal or not, let alone know the tolerance for that squeeze.

Mabye this was true in the era of Quake 1, but not anymore. 3D games have evolved since then, and are much better - either through graphics or some other complaint based on the difference between 2D and 3D.

The reviewers whining about this sort of graphics is just superficial. The real quality of the game is not how it appears on screen, unless there are glearingly major problems that interfere with gameplay (either through obscuring critical information, showing information that should be hidden, or by being distracting).

The artwork.

[Sludge]

The artwork on some of the old Nintendo games was amazing. The established art style in SNES Zelda is a perfect example of expressive simplicity.

I love oldschool 2D games, and I think there is still a place for some of that in the indie scene. You need really good artists to make it worthwhile, though.

In fact, if you know of one, contact me. I may be able to offer a job.

Yes, I agree

[Tom7]

I feel the same way. However, I think that I finally do see the light at the end of the tunnel. An artist working in 320x200 VGA space can certainly make careful use of pixels to provide a much better visual result than one gets from the Quake I engine. But as resolutions increase, the difference becomes much smaller: an artist working at high resolution is essentially working with vectors rather than pixels (think of a painter or cartoonist), and so in some sense has already lost his "pixel perfect" advantage.

It also seems that we are getting so much 3D power recently that it's no longer enough to simply have dazzling numbers of alpha-transparent triangles. 3D games are needing to resort to more interesting visual styles (cf Zelda: Wind Walker), and I think that may ultimately bring them to the same artistic levels that we see in modern "2D" games like the Capcom fighters or GBA side-scrollers.

Re:It's a matter of time.

[ultramk]

Sure, there has been sculpture too, but never in the sheer volume of traditional two dimensional art.

To start with, full disclosure, IIAS (i am a sculptor).

Uh, I call bullshit. The earliest examples of art we have are not cave paintings, but small carved figures. Lots of them. There is no way on god's green earth you can make a blanket-statement like "2D is more common than 3D". Show me an ethnographic study of the world's cultures (historical, too) that proves that paintings or drawings are more common than sculpture, and I'll eat my hat. (I have spent far too much of my life in art history and anthropology classes, so I feel confident that my hat will remain uneaten)

We live in a 3D world. 3D visualization is inherent in our genome.

2D visualization is not inherent: there have been studies of some isolated cultures with no 2D representative art, and if you're not introduced to the concept of a 2D image by a certain age, you just won't "get" it. Look at the representative art of the Australian aborigines, where every painting is in the style of an exploded anatomical view. Why? Because the damned dingo HAS 4 legs, damn it, so you have to show all of them. None of this side-view perspective crap.

To get back on topic:

The reason the tide of public opinion on this board prefers 2D games to 3D games is because that's what you all grew up with. It's what you cut your teeth on, and it's what you prefer. To sum up, if you honestly think the crappy games of our youth are better than today's games, YOU ARE AN OLD FOGY. I grew up at the same time, but I have no illusions as to how much most of those games sucked. Y'all have very selective memories and are the modern equivalent of those people in the 80s who refused to acknowledge that video games were better than pinball games.

Just my opinion,
m-