I’m a dedicated artist, transitioning from 2D to 3D CG art. I work with Maya and Photoshop primarily and have done mostly just static meshes in the past, some rudimentary animations. I’m currently taking an animation class for maya so hopefully that will not be an issue after this term… Yes, I am a game design and art student but I want to learn more outside of the classroom when I can.
I really want to learn how to create 3D Art Assets for Unity specifically. I’m not looking to learn scripting for Unity as other teammates in the project I’m in are working on that. Scripting’s not really the focus of what I expect to do as a 3D artist (maybe this is naive? ). I hope to build nice assets for an environmental art and character art portfolio and learn the ropes currently.
Most (if not all) of the tutorials I’ve seen for Unity are specific to scripting in C# in unity, but not really focused toward creating 3D assets. Are there good tutorials for establishing a 3D modeling pipeline for Unity which can be recommended? Can any experienced 3D art asset creators give me some tips as I start out?
The pipeline to Unity is pretty simple compared to some other game engines. 99% of the work is done prior to Import from Maya or Max (or whatever you may use).
Instead of trying to figure out a Maya2Unity pipeline, I suggest you concentrate on making game models/textures in general. Until they’re exported for use in a real-time engine, the standards are pretty much the same. You’re not going to find many (if any) tutorials for making Unity models because that’s too specific. Any game modeling tutorial will hold true no matter the game engine. Only the import/setup process varies by engine for the most part and sometimes engine-specific materials.
I can see how this is true, with the base model being more or less interchangeable while being worked on in Maya despite the engine it ends up in. This is why there are so many hair models from the sims which keep showing up in mods for other games like oblivion. (har har) I imagine the polycounts are also pretty standard across engines, though I’d rather ask than assume if you know what I mean.
I’ve watched some of the BergZerg tutorials such as the ones pertaining to trees in unity, which is less similar to other engines I’ve used in the past. Are there other facets such as that which I should be aware of? Clearly terrain deformation will work differently with Unity also, and no doubt there are tutorials for that in BergZerg’s list as well…
I’m not familiar with the BergZerg tutorials. Unity has a built in Tree system that is procedural and they react to wind effects and have a built-in optimization function sort of like LOD where the mesh trees are shown at one distance and a very low detail billboard is shown at farther distances. These trees are used with the Unity terrain. Of course you can model your own trees and terrain in Maya and import into Unity, but they won’t have the dynamic features of Unity wind or the ability to use multiple terrain images “painted” on (well you can if you use a splat shader on the mesh, but that’s another subject).
There is always a debate about polygons in any engine. For the most part modern engines can handle way more faces than you’d want to use. Keep in mind that face-count is only one part of optimizing. Other than code optimization, you have to think about textures memory, dynamic lights, image effects (full screen), texture effects (like render-to-texture for mirrors, video screens, etc.), physics, AI, etc.
Your mission as an artist is to make an asset look as good as possible keeping in mind the balance required for an optimized game world.
The Unity trees and terrain are not any more low poly than something you might model, but they are updated dynamically and adjust polygon counts and/or LOD stages automatically (this is my simplified explanation) depending on your settings.
Lots of important points there which get me thinking, thanks! So far my job as an artist has been largely to fetch coffee and watch tutorials while I wait for the others to get back to me on the concepting drafts I’ve submitted. That and being team cheerleader. It’s honestly refreshing to getting back to thinking about what I will need to do for -real- work.
I’m really impressed by what I’ve seen for the trees alone, as they’re based on the Fibonacci sequence so they get a lot of naturalistic variation going on. One thing I noticed was in the BergZerg tutorial for trees in Unity, he was getting a lot of slowdown while making trees. Is this fairly normal, and something I should be aware of while working? I’m curious to know whether they bog down processing if there are too many complex trees, though I bet that’s a factor of LOD settings and what platform the game will be viewed on.
I’m going through a mess of tutorials now, trying to pick out the art/maya related ones from all the scripting business end of things.
It depends on how many trees you’re working with, the detail level of each tree, the speed of your computer, what else is going on in the Editor at the time, Editor quality settings, etc. And when creating a video tutorial, there’s another fairly large hit on the video/CPU memory while recording.
3D assets for games are pretty standard for the most part, there is no “Unity way” of creating assets. I can create an environment that can be used in CryEngine, Unreal, or Unity- doesn’t matter. There are some technical aspects that are propietary to engines, for example when getting geometry into CryEngine, you need to use Crytek’s Cryexport tools and not FBX like with Unity, but these are just some technical data exchange principles. A good game asset (i.e. good topology, fair detail where needed, optimized mesh, great textures, maps, etc.) will work in pretty much any engine.
I’ll be sure to keep an eye on those factors while working with trees, thanks! I had some slowdown while recording UDK level walk-throughs with camtasia, so it would be fair to assume the same to be true with Unity.
Awesome, I appreciate the link and will look through those tutorials. I’ve mostly worked in Maya for rendering and not exporting with the courses I’ve taken, which was my primary concern. It’s that step of exporting a model which works well (in any engine) that I want to really nail. Thanks again for the link!
Yeah data asset exchange can be tough when you don’t have the technical know how or experience. Fortunately, that’s one of Unity’s strengths. The asset workflow is one of the best in the business, and this is coming from someone with experience working with Unreal and CryEngine. It’s as simple as creating your assets, whether it’s rigged and animated characters, or static geometry/environment props, and then selecting and exporting out of Maya/3dsMax/Softimage and importing the FBX files into Unity. Usually everything works. There are some small caveats that need to be taken into consideration depending on which dcc package you’re using in your pipe, but the whole process is about as easy as any game engine has ever made it. With all the game engines I’ve worked with, Unity is the easiest and most predictable I have used so far.
I’m excited now to see how it goes after watching a lot of these tutorials. Sure, learning it will be daunting but with the replies I’ve had here I’m relatively optimistic! Thanks for your insight on this, it helps my post-quake malaise as I’m getting back into tutorials. XD
). I hope to build nice assets for an environmental art and character art portfolio and learn the ropes currently.