Is 2d really "expensive"?

[rainkaimaramon]

I was doing some "research" about 2d games, mainly how to convert hand drawn stuff into detailed what nots; that's an entire thing all together, when I notice a lot of people on game dev, dev this and dev that say it's more expensive to make 2d games do to one factor or another, I wasn't paying attention usually the whole thing spins out of into business talks that sound more and more like a potzi scheme then actual game development.

Anyway it is often repeated that 2d games is too expensive to make even though I look around on the net and hundreds, thousands even, of pretty good looking FREE 2d games and many of them are playable too... Ahem.. Which leads me to the question of why, why are these drawn (many times pixeled other times not) more expensive then hiring an actor, using a lot of equipment to scan and map said actor, having teams of modelers, texture artists and art directors do touch up, skin, rig, pipeline and render the thing then a few guys drawing a few frames, scanning, clean up and getting it out the door? I guess it's a number game, or something.

PS: Before anyone points it out I know making things in HD at 90 frames per second is hard, but my stance is why do you need that? SD at 24 (8x3) frames is fine by me, just saying.

[Calin Leafshade]

It's about economies of scale.

Setting up a 3D workflow does take some time. It certainly takes longer to model a 3D character than draw a 2D one but animating a 3D model is orders of magnitude faster than 2D. So if you have a large game 3D wins out in the long run.

[Igor Hardy]

Anyway it is often repeated that 2d games is too expensive to make

Nope. You have apparently asked around in the wrong places or confused "hand-drawn" with 2D which is not the same. 2D vector graphics are the cheapest and fastest type of graphics to do and animate (and they easily convert to any res you want). But if you don't need to animate, even a pretty detailed hand-drawn background can be done far easier and faster than a comparably decent looking static 3d location (especially if you want the player to be able to move around everywhere, look at the 3D objects from every side and angle etc.).

[SpacePirateCaine]

The issue is measures of quality, reusability, scale and workload. 3D takes a lot of time on the front end of the job in the actual creation of individual models, drawing textures and what-have-you, but once a single 3D model is created, the same model can be used for all individual animations. If the team involved knows their stuff, they'll be able to rig a set of animations that apply to all individual models, so each animation also only needs to be created once.

If the engine is designed for it, it's even easy enough to make all character faces generate procedurally from a set number of variables, creating a huge (Nearly unlimited, for the purposes of game creation) variety of character heads, usually on a relatively limited set of bodies (BioWare and Bethesda make prodigious use of this). This further reduces the cost in man-hours.

Once you have a mo-cap studio it also becomes somewhat easy to animate - and at least in my current company, we have people whose specific job it is to be able to act out most necessary movements for any given game, as well as understand animation rigging and implementation of animation format within a game. At the end of the day, assuming that a studio is using animation economically, you can have hundreds of characters animating in hundreds of different ways and the workload is much smaller than, in the case of a 2D game, where each 2D frame must be created newly.

There are a lot of shortcuts, of course, as Ascovel pointed out. Vector graphics, and the huge trend lately in 'paper doll' animation (Individual parts pinned at joints) has helped to keep 2D viable, but in today's game market where the average consumer in the more profitable demographics expect fully-interactive environments with working physics and the illusion of variety, 3D is the go-to solution.

2D of course has its benefits - I prefer it far more than 3D when I'm developing on my own time and dime, but from a business standpoint, if you want to go big, you're usually best off going 3D.

[Igor Hardy]

There are a lot of shortcuts, of course, as Ascovel pointed out. Vector graphics, and the huge trend lately in 'paper doll' animation (Individual parts pinned at joints) has helped to keep 2D viable, but in today's game market where the average consumer in the more profitable demographics expect fully-interactive environments with working physics and the illusion of variety, 3D is the go-to solution.

2D of course has its benefits - I prefer it far more than 3D when I'm developing on my own time and dime, but from a business standpoint, if you want to go big, you're usually best off going 3D.

Leaving aside the big business standpoint, everything you mentioned is possible to do in 2D vector graphics - only you can do it quicker, cheaper and with a smaller crew.

Even modern motion capture should be possible to quickly translate to a flat image, if one was so keen on using it. But you probably won't motion capture animals, vehicles, explosions etc., so super-realistic 3D animation has its challenges and it's still pretty slow to realize, while simplified animation and flexible object shapes are very useful for presenting out of ordinary situations with much less hassle.

3D brings the mass audience - there's no denying it. It's an entirely separate issue though.

[SpacePirateCaine]

Absolutely right on that one - I was speaking primarily from the big business standpoint, since that is almost definitely the context from which Rainkaimaramon was speaking (Developers discussing the benefits/drawbacks of 2/3D in professional development). Working on the independent level, or on mobile/small-scale platforms leaves a lot more room for choice in the matter - you should consider what your project really is, and whether it would benefit from 3D or 2D more. Most adventure games, particularly on the scale that we as a community work on, on average are at the scope that 3D or 2D wouldn't change much at all, cost-wise, if 2D wasn't drastically cheaper.

Now, for a first-person or over-the-shoulder third-person game, 3D is almost a necessity. But that's sort of a moot point in this conversation, I would imagine. For the sake of this argument, we should probably stick to games that can be done in 2D as well as 3D, like adventure/puzzle games. Ascovel, do you have any good examples of games that use 2D vector graphics that are still as richly animated/maintain the visual quality of a hand-animated 2D game? I'd like to take a look, and I imagine the OP would enjoy some context as well.

[AJA]

Even modern motion capture should be possible to quickly translate to a flat image, if one was so keen on using it.

Yep.

[Igor Hardy]

Ascovel, do you have any good examples of games that use 2D vector graphics that are still as richly animated/maintain the visual quality of a hand-animated 2D game? I'd like to take a look, and I imagine the OP would enjoy some context as well.

While you can't match certain aspects of hand-drawn animation with vector animation, what about such popular and visually dynamic indie games as World of Goo, Limbo, or something as outrageous as No Time to Explain?

EDIT: Oh, and it looks like the Ubisoft funded Rayman is going back to 2D for the series' newest installment: http://www.youtube.com/watch?v=YmsD94FFOoQ

[SpacePirateCaine]

Great examples Ascovel. As I mentioned before, the shortcuts by and large are reliant on single static images rotated around at joints in the 'paper doll' style that I had mentioned before. Animations are a lot more jerky and can appear a bit 'cheap' - I had noticed this in Limbo as well - though at the end of the day, most people don't have a problem with these slight graphical idiosyncrasies.

This is definitely a good way to cut back on your animation budget, however. If you want the animation to appear smoother or more 'realistic', it requires redrawing more frames, which quickly will eat away at costs, so it's a matter of finding the middle ground that works best for you. Some studios are also embracing the technique and designing their games visually in a similar way to accomodate these issues, such as Grasshopper's Black Knight Sword.

[Anian]

Indeed, if you don't have budget/talent/time for the best then take your disadvantage and turn it into "style".
By that I don't mean "I think perspective is too time consuming so I'm gonna do as I like" line of thinking, but lets take old HannaBarbera cartoons, like Fred Flinstone (and every other character) having a head separated from the body so the production and time cost goes down (you can check this video, that's just what thought of since I've seen it recently LINK)

Btw if you really have to ask this question, I'd assume you haven't actually tried to animate a 3d or a 2d character, try it and see.
Basically both solutions (3d models and vector art) reuqire a lot of preparation but then you can make stuff quickly. Hand drawing and pixel art (for the most part) require more effort but are somewhat easier on the preaparation part - here's a blunt example, to get the point across - so let's you have a tv cartoon series, then it's "easier" to go with 3d or 2d vector art because you make production once and then you reuse assets (hands, faces, walking animation etc.), but if you want to make like a one short cartoon, then you'd go with hand drawn...again this is a overly simplified example.

...and just a big WTF for 90fps. The movie industry is thinking of going from 24/25 fps to 48. and that's a big leap. Also it takes about a day or two to render (3d art) a frame from a Transformers movie, so you'd be working for 3 months for 1 second of your movie. Let's say youmanage to render a whole second of animation per day, it'll still take you months to render a short cartoon...