How to do research and development for a new art style
Suppose you want to figure out a new art style for a game project. The big questions are how to develop an appealing style that fits a certain theme, and also provides a reasonable production workflow for small team.
Is the quickest way to find that style through working with a concept artist? I suppose a concept artist can help you visualize themes, and then you'd have to do a little work translating each type of asset into 3d to confirm your production pipeline?
Is that the basic idea more experienced teams are doing? If you wanted to work with absolute minimal team, what specific skills are you looking for?
More specifically, suppose you want the feel of a gritty apocalypse game - similar theme to The Last of Us, for instance - but realistic graphics on large scale wouldn't be feasible. Any examples you can point to for simplified style that still captures same feeling?
Replies
one thing i did when ( together with Luis Antonio on 12 Minutes ) i was developing the artstyle for the game, was to be inspired by classic paintings, in this case, impressionism, wich i tried thru several 3d iterations convey the feel to the art style, some i exagerated, others i tried to be more moot. it was a fun process, ill try to find some images of the work in progress until we hit a style we were happy with.
I am out of the loop with current games but I looked up 12 minutes and it looks fantastic and very interesting. Nice to see something unique and thoughtful.
About developing the art style, did you follow any kind of process, like have concept artist create many thumbnails and kinda storyboard the gameplay loop first? Or build a few 3d scenes first? Any best practices you could recommend to avoid time sinks for a process of art R&D?
some times what you need is... time sinks :)
to make something that stands out, you will have to test things and play with them, iterate on them and scrap stuff that doesnt work.
personally i would suggest to work with concept artists / 2d designers on it, you can design stuff in 3d and come up with cool styles but usually 2d guys are just quicker and push out a lot more content than you could in 3d. and even if you have something you like in 2d you first will have to make it work in 3d and iterate on it together with the 2d artists, find solutions in 3d that are easy to replicate, maybe even automate to cut down times.
if a stylized last of us was my goal, i would first of all take a screenshot from last of us and try to stylize it, work on proportions simplifying texture detail etc.
before Ori and the Blind Forest became the game by that title, it used to look like this
then it went a whole lot simpler to nail gameplay
it went through a lot of stages before it became the final vision of the game.
On top of what Neox is saying about iterating and scrapping I would say it's also worth trying to find projects that were made with similar team sizes that you intend to be working with and seeing what they were able to achieve. Referencing older generation consoles can also be a good way to create an aesthetic that plays to the "retro" appeal while cutting down on production time. A good example would be Pathogen-X on itch.io.
Also figuring out how much shader/materal/lighting work do you want to have to rely on or if you can achieve your goal with lowpoly/diffuse and just brute force the visual style. I've been referencing a lot of 2D games for a 3D project because they're so successful at doing so much with so little such as Hyper Light Drifter and Titan Souls.
A visual style like Death's Door had 6 main people on the credits, with only 1 full time modeler so their style was very much dictated by their available resources. A good interview with them on youtube.
While I can only offer an opinion on it from what I've been able to gather from researching this is a lot of incredible and unique visual styles have come from working within limitations; Whether it's the tech or the resources/time. So figuring out what your limitations are and figuring out how to work within those can produce some really awesome results.
Great topic, for me I like to use Pinterest, Then PureRef the best images from your Pinterest and complement them with ArtStation and Maybe a quick look on Google/DeviantArt, some of your own photos. Then use that as an inspirational starting point for your Concept Art.
Also Atomhawk have kindly put together six videos on there visual development pipeline for “Project Canopy”.
https://magazine.artstation.com/2022/01/announcing-studioquest/?utm_source=homepage&utm_medium=toprow&utm_campaign=2223-new-learning-series
Thanks guys, lots of useful stuff here to think about.
I'm hoping to start this work a little later this year, now I got a better idea how to approach it. I think I'll do as much prep work on my own by doing paint-overs and such from inspirational art, then if I am having trouble settling on something solid that I can also translate easily to 3d I'll see about getting help from an expert.
I do like the idea of sticking with a rudimentary 3d style, like limit materials to only albedo and just go as far as I can with basic art fundamentals to get it looking nice. Seems like that would focus my art time more productively. And there is games from late 2000's that still hold up very well today and in some ways I think they are nicer to look at than some of the hyper-detailed modern stuff coming out recently.
One starting step has been setting everything like it should be on the final product, technically wise, like if you know you can't go with high geometry, high material complexity etc, set it to as optimal as it can be even if it looks like a programmer game, and then from there see how you can manage to reach the visuals you want without impacting performance
I think that will be especially important as I want to write characters and story for this game. Even in my current project which is just pure simulation, one of the major lessons learned was i should have setup the game framework first before doing any of the art. Of course I had to figure out a lot of stuff as I went but def going forward I'll be setting up the "bones" first before doing any custom art.
Ideas are worthless. Inspirational art is worthless. Play to your strengths. If your artists know hentai tamagochi then that is what you do. You do not figure out what to do, you do not follow your vision - you just do what you already know best. Don't make it harder than it needs to be.
These guys who have done worthy work that I can see for myself are suggesting otherwise.
Unfortunately I don't "have artist." I am a mostly solo indie developer, I don't have a unique specialty, but I do like for my games to look competent, memorable, and also be something that I can finish in a reasonable time.
Usually I just go with realism because it is easier to get assets for that style, and it sort of suits my interest anyways. But I think for the type of game I described, realism might make the scope too large. Hence, wanting to do some R&D to find a new style.
Forget "guys". Use brain. If you have well-developed skills, then what you do with those skills is going to turn out well. If you intead go down the path of "vision" and enter unknown territory there is no other option than to aquire new skills to adhere to vision, which requires investment and involves uncertainty. These are facts, not opinions. The facts stand regardless of whatever you believe your vaguely-referred-to "worthies" seem to you to suggest.
Is it not immediately apparent to you that whatever you are best at is what you are best at? Is it not immediately apparent to you that it is better to play to your strengths rather than your weaknesses?
What are your strengths? What are your passions? That is where you go.
Neox has pretty much nailed it
You need to go round in circles a bunch of times - the real skill is in learning to iterate quickly on ideas.
you absolutely need to get a gameplay template working before you go anywhere near lookdev because gameplay should be one of the drivers behind art style.
the gameplay template and basic premise will inform on critical path art features - eg. you can't make a fishing game without defining what water and fish look like.
the rest of the look can grow from the critical path art features and be guided by how the game plays
This is a weird take, to me. Like, if you're fine making a mediocre project, in order to just "get something out there", then sure, this is probably good advice. But if you want your game to mean something, to innovate, to make an impact beyond "I have five minutes to kill, I'll spend it playing this game", then that probably means looking beyond yourself, the way most good art does. It probably means pushing yourself to go down rabbit holes, to explore, to investigate—
Most games I really love are people following their vision, people who have ideas about what their ideal game would be, and don't see it out there in the world already. People who go vastly out of their comfort zones to accomplish something they are truly proud of.
Don't get me wrong, I agree with some of your comments; ideas without action will not make a game. All the inspirational art in the world won't mean anything if it doesn't spur you on to create. Similarly, you do have strengths and passions you must play to, if you don't want to fizzle out.
But that's a lot different than "do not follow your vision" or "do what you already know"; which frankly, sounds to me like a recipe for a very mediocre and mundane clone of a game. It's not what will push this medium forward, or fulfill you.
Yeah I am all about doing the most pragmatic, business-minded thing, but it must balance with fact that I can't just make things with only goal to make money and care about nothing else. Only make something that panders to my notions about what an audience wants. To me that is like pretending to be a persons friend but secretly you hate them. I can't live with myself that way. I can only spend time making things that I really want to play myself.
It is so much effort to make a game on your own I doubt anybody could complete one without such an attitude. Only with a team that does all the actual work do I think you can divorce yourself from the product in such a way.
whut?
I get your point. Working with your strengths is important and knowing what style lies within your capabilities. But how does that make ideas and inspirational art worhtless? It doesn't.
Inspiration and a "vision" is the key to stunning and unique gameplay experience. Obviously, it should be within your teams capabilities and also should use their individual strengths for good. But that is exactly what this thread is about: Finding the inspiration, the style, the overall vision of a project that is likely to be accomplished succesfully.
About finding inspiration and a vision.
I would sum this up as a process which evolves from a state of mind, a feeling really, to a fleshed out idea with a clear vision. And I think this process applies to all kinds of art, including music or writing.
I think a good idea already includes a direction for the art style of the game. The core gameplay already defines what fits well. On the other hand if you have an idea and then just pick any style that you like (found on artstation) and kind of fits the idea, it could turn out to be unauthentic. To me, it goes hand in hand: the idea for the game and the art style that it should be amplified with. Amplified, because the art style and overall feeling is what drives the games authenticity and how easily the player can be immersed in the game.
This should be your main goal when trying to find the right style. A style the amplifies the core idea of your game, that represents it in the most authentic way possible and draws the players into your game world.
Context. He wants the "quickest way" to establish an art style, for an "absolute minimal team". The quickest way to establish an art style for a minimal team is to utilize the existing skills of the members of that team. For those who believe that will only lead to mediocre results; why? The team chose to become artists in the first place because they were inspired by other art. And if, according to you, that inspiration led to mediocre results, why should consuming more inspirational art somehow lead to anything other than more mediocre results?
It would typically be far more beneficial to research technology. What can be done with generated/modular/procedural art? How about using medium poly meshes instead of normal maps? Or voxels? How about making the setting something like planet Cybertron so one can use those those popular Blender greeble extensions to efficiently create detailed sci-fi meshes? Or basing the whole thing around existing paid asset collections and making variations on those? How about keeping assets extremely simple but get visual quality out of ray tracing or effects? I could go on an on with feasible strategies, but that's what they are - strategies. They require technical knowledge or at least creative thinking to pull off, not just the rush of inspiration one gets from looking at awesome art. Once you come down off the high you realize someone actually has to do all the work. How to get the work done should be the top priority.
So, what are your strengths? And then, how can you be clever about utilizing those strengths in a way that maximizes your efficiency? Can you make new tools to help you or use existing ones in a novel fashion? Ideas are worthless unless you can actually execute them and what you can execute is defined by your existing skillset. If you just work on it, and if you push yourself, you are bound to find a smart way to approach this.