2d vs 3d: theoretical costs all things equal

Hi, I am a hobbyist. I am looking for a long term project over the next 2-3 years to complete with a friend. Yes, I do make smaller games to hone my skills. I have a year of programming under my belt with online classes and self-teaching. Programming can be stressful but it is good, wholesome stress for me and I find challenging periods ‘fun.’ However, I frequently mess up the simplest of stick figure drawings and the thought of having to create art assets is simply terrifying.

Due to this, I plan on hiring out the art asset work after I show myself I am serious about creating this game over the next few months. The game is going to be a turn-based game similar to Heroes of Might and Magic or early CIV games. I see no reason it needs to be 3D, however, my friend working with me believes my assumption needs to be challened and I accept that he may be correct. So I turn to you and your endless years of experience to answer what is likely a difficult question conceptually.

Caveats: I am NOT asking about the coding. I understand Unity is amazing and does plenty of the heavy lifting 2D or 3D. This is about art assets and the difficulty/cost of their creation. These are theoretical questions to gauge if there is an intrinsically higher amount of work in the creation of one form of art over another.

Questions:

  1. Theoretically speaking, if I found someone with artistic abilities such that making sprites or 3D rigs/textures/etc were in complete and total parity in terms of skill…which would cost more money? Same place geographically, same currency exchange, etc.

  2. Theoretically speaking, if I were to split my body into two parallel-running universes, one locking himself into a large steel cage with a computer and some great 2D asset creation program for a year and the other locked into a similar steel cage with a computer and some great 3D asset creation program for a year and they spent exactly the same amount of time learning the programs (having no prior experience in either 2D or 3D animation and no difference in natural ability for one over the other)…at the end of the year and the cages unlocked, which artist will walk out more readily prepared to create game-ready assets?

It goes like this.

It is impossible to answer your question #2, because the answer depends on your natural inclinations. If you are more predisposed to drawing or modeling, corresponding discipline will look better.

The answer on #1 will depend on project type. Basically, a 3d game has advantage of freely allowing you to display same character from a different angle. However, a 2d drawing is more flexible, achieving “fluidity” or expressive characters is significantly easier. Meaning… if you have a character that needs to be displayed from different angles, 2d approach most likely will cost more, ebcause you’ll need to portray same move from different angles, and this will need to be drawn. However, if you’re trying to create something that looks like 1930s animation (see cuphead, for example)
For example:

Then this is insanely difficult to achieve in 3d animation and will require a very highly skilled artist or an art team. 3d animation and assets run a higher risk of looking wooden and robotic than 2d drawings.

The other interesting thing is that 2d for some reason tends to age better than 3d.

Basically, it goes like this:
A cost of individual 2d frame is going to be lower than a cost of an individual 3d character. Meaning too many frames - and 2d character will cost more. However, 3d character comes with possibility of controlling its movement via scripts (ragdolls, inverse kinematics), free ability to attach objects to it (hats, clothes, etc) and recolor them or change their material. However, in 2d it is significantly easier to achieve highly expressive images than in 3d. As someone said “I have seen many times when an amazing 2d concept art turns into a mediocre or bad 3d models”. A 2d stick figure can be made more visually stunning than a 3d model.

Basically, there’s no real clear cut answer there, and the answer depends on your project, its art style and goals (pixelart may be cheaper than 3d model, for example). Depending on what you’re trying to do, either approach may be a better choice.

5 Likes

I agree with your friend that 3D may be the better route, unless you’re specifically trying for an 80’s/90’s feel to the game. The Civ series starting with Civ4 really improved its look by going with 3D models for the game world.

Your questions are virtually impossible to answer though, as they both really depend on your natural talents and what type and amount of art you’re really going for. As a side note, you can get a long way with 3rd party ready made models for your game if you go 3D, but finding already done 2D art is far more difficult.

So this is not really related to either of your questions, but i think it’s still important to note:

If you keep things in 3D, you will have more flexibility down the road. It doesn’t matter whether you have created a 3D character by yourself, or if you bought one. As long as it is rigged correctly, you can always add animations to it. You don’t even need any artistic skills to do that, which is certainly a plus if you ask me.
An asset like FinalIK can add additional flexibility to your character, without requiring you to even leave the Unity editor.

In 2D, on the other hand, you will will be restricted in many ways. If you didn’t draw it yourself, it’s nearly impossible to add anything to it without looking weird. The perspective is fixed and has to look identical on all your graphics and you are limited in the look&feel, because you can not play with lighting like you can in 3D.

So the advice from a professional programmer that has tried to do all the art stuff for years…go with 3D or go with 2D pixelart (top-down/sidescroller).

2 Likes

Just thought I should add: you can make a 2d game using 3d modelling to create the sprites, and you can also feature 3d models in an otherwise 2d game.

Ultimately it doesn’t matter what anyone here says, it’s got to be what your friend wants to work on - trying to convince them to make a game they aren’t passionate about or don’t see working is doomed to fail.

4 Likes

Having been involved in contracting for indie 2D and 3D art, 2D art is generally going to be less expensive. It’s easier to cross over from traditional media to 2D than to 3D (which is a more specialized skill). The larger pool of 2D artists drives prices down. Apart from characters, 2D environments generally cost less, too. Regarding characters, depending on your game design and artists, you could reduce costs by using skeletal animation rather than spritesheets. Ultimately, the points made above, especially about what both of you want to work on, as well as your design and technical abilities, will probably have more weight than the upfront cost of 2D vs. 3D.

4 Likes

Exactly what @TonyLi said, it is completely dependant on the type of game… For a 3D space side scrolling shmup the level of difficulty and time investment is less than many games including a lot of 2D titles, when you start evolving into a 3D openworld RPG vs. a SNES style 2D RPG then the amount of skill / time / money / investment becomes massive on the 3D side…

Design decisions are truly where the impact lies…

1 Like

Should it matter this early in development, you could prototype/asset store/open source both 2d and 3d art assets for your game and trial both approaches or keep both approaches open and develop your game.

Unless you have very low entry level specifications (e.g. WebGL, Low Spec Mobile) or are going for a retro look I would opt for 3d but that’s my opinion.

I would argue that it’s almost more about art direction and look n feel than the type of game, given the current level of gaming hardware and Unity game engine.

Maybe a mix of both would even give you the look your after with minimum performance issues.

I agree with all above.

It’s worth mentioning though that Unity is a 3D engine so because of that - and the fact you will be creating a game that itself benefits from orthographic camera projection - I’d say 3D is the better option for the type of game you are going for.
The reason for this is pretty basic but can save you some quality time in development - with 2D the point of view for the camera will rely on the developer creating some “rules” for sorting based on Y value and a lot of edge cases for these sorting rules, which could simply just work out of the box naturally in a 3D orthographic (z depth) setup.

It is also worth noting 2D assets can be used within a 3D world, though I think there are still some sorting issues with this set up and lighting/shadows come into the equation - which equals more time spent on solving issues that are naturally solved and “just work” with 3D assets/content.

Granted - I have not worked on a game like this before so any developers with more experience on this subject can correct any of my incorrect assumptions - if I’m wrong about them.

Good point a 2D isometric world needs dedicated coding, tile logic to ensure walls and tall features appear above their background tiles, you could do 2D isometric with 3D tiles and get the sorting done for you and gain the benefits of 3D lighting (something else you would need to code).


I’m pretty sure the dungeon in this game is 3D with 2D sprites with normal maps that work with the lighting.

Devs Blog http://robotloveskitty.tumblr.com/post/33164532086/legend-of-dungeon-dynamic-lighting-on-sprites

https://www.youtube.com/watch?v=_NyoyOc1qWk

1 Like

I don’t need to use theory. 3D is more expensive, by far unless you go right down to PS1 / N64 detail levels. This is because 2D can get way cheaper with retro and require very, very few frames to tell the story and still look effective vs 3D.

What you should be looking at is doing a prototype. When you have that, you can ignore the quality side of things and just use 3D or 2D placeholder stuff. This allows you to do the game and worry about the art right at the very end. Risk is substantially reduced, and you can swap between 3D or 2D as your situation dictates when the time comes instead of spending up front which is tbh foolish. You said your game works with either style so it shouldn’t matter to use programmer art until such a time arrives :slight_smile:

Many a developer realised a bit of post effects and some ingenuity made their games look absolutely fine with programmer art!

Do prototype. It answers your own questions.

1 Like