Trending
Opinion: How will Project 2025 impact game developers?
The Heritage Foundation's manifesto for the possible next administration could do great harm to many, including large portions of the game development community.
Featured Blog | This community-written post highlights the best of what the game industry has to offer. Read more like it on the Game Developer Blogs or learn how to Submit Your Own Blog Post
This article is for background art production of 2D games. It deals with 2D & 3D assets management. It would help you save troubles when you need to edit, and reuse the graphics.
Today many types of games use the same kind of illustration for level background: slot games, Match-3 games, adventure games, etc. Just like this example.
In this article, I would discuss how to manage assets for background art creation. The purpose is to get things in order and save production time and budgets.
In producing background art, our team use 2D or 3D pre-render solution for different projects, or mixed solution. So we have both 2D and 3D assets to manage.
Please note, we use 3D assets to construct the scenes, and render flat 2D images to use in the game. We don’t make 3D scenes for 3D engine games, so we would mange 3D assets differently.
2D stock photos could be put in three categories:
Texture files: these are useful both for 2D and 3D art production.
Photo objects: these graphics are often put as a whole into your picture, and then have some overpainting added to blend into the scene. Usually useful for making interactive and hidden objects.
Background photos: we often cut part of these type of photos to make part of our scene.
In our studio, we store these three categories of photos in three libraries. And each library has its sub category. I think every one is familiar with this hierarchy.
Of course when an artists looks for a picture, she doesn’t go down from root to top. Instead she would use key word search. So, each picture and the its folder must have relevant file name. Use your local language for the file names, for not everyone in the team is good at English.
For 3D models, we have two general libraries. The first library is a hierarchy libraries, pretty much like the photo object library I mentioned above. It has 3D models of often used items: tools, furniture, kitchen utensils, cars, etc.
In doing every scene, our 3D artists would download some fresh 3D models from the web, which they can’t find in our own library. After that, we would put those newly downloaded 3D models to our library. Over time we would have a comprehensive library for all the things.
And we arrange files in this manner: one 3D model file for a certain object, for example a chair and an umbrella. And each file has an eponymous jpg file with it, which is the screenshot of this item.
Think about this scenario, when a 2D artists looks for a chair of particular style for her scene , she opens the folder in the library, and see 20 chairs in the folder: Chair 01.obj, Chair 02.obj….Chair 20.obj. Having no way to preview these chairs, she would have to import these chairs to the 3D scene, that would take a bit time, for each chair is a high polygon model. But if we have jpg files go with these 3D models, she can take a quick look over the thumbnails to see which is the desirable one.
And the second 3D model library is arranged by projects. For example:
That 3D artists can look back into the previous scenes to find re-useable assets.
For a particular project, it has a theme, that in fact we would intensively use 3D models and textures from a narrow range. For example if a game features the inside of a medieval castle, then the most often used 3D models are: furniture, sculptures, weapons; Textures: wood, stone, fabric, metal, patterns. We would use a small project specific library for quick finding assets.
In general, although an unflattened PSD file would give you lots of freedom when you need to tweak the picture, the final PSDs we keep are with layers merged to certain extent, that the file is smaller and the layer structure is much more clear.
Generally speaking, on the background picture, we give each separate object its own layer: a tree, a chest, a house. Often one layer for the object itself and one layer for its shadow. Yes in painting the object, we have lots of layers, for strokes, highlights, layer effects. We have them all merged later, to make the file easier to manage.
For objects need multiple states, one layer for one state.
For objects with simple animations, one layer for one frame.
In general, for a scene we would have a 3D file, with all the stuffs in it: 3D models, lights, cameras; And a bunch of textures.
We want the 3D files well organized, that they are easy to re-use. If you don’t make set rules, that different 3D artists would give you 3D files organized in different ways, often poorly organized.
We mostly use Maya, I think for all 3D applications, the rules of making a clear file are the same.
Firstly, for objects, always have a single merged mesh for an object, and give it a relevant name:
Don’t leave useless meshes or empty objects in the scene:
And put things in relevant object layers to make scene easy to manage:
And we often render the following images out from a scene: a regular image(beauty pass), an ambient occlusion image and some images serve as object masks. We can setup render layers for this purpose. For all the 3D scene files, we use the same kind of setup.
So all the 3D scene files are organized in the same way, that in the future if a 3D artists would go back to a certain scene, she wouldn’t need time to figure out what is for what.
And all the textures for a scene should be packed together, going with the 3D file, to avoid losing textures.
And one last word, all the source files and final files for all the projects should be backed up with good organization. This involves version control and file backup solutions, there’re many out there, you should choose one works for you. And sorry I wouldn’t discuss it here, it’s not the main purpose of this article.
Read more about:
Featured BlogsYou May Also Like