Robert Tamayo


Thoughts on Making Games with Unreal and Blender

After a very short amount of time working in Blender and Unreal, I had a character running and jumping around the demo level. By that point, I hadn't written a single line of code. It's actually quite incredible what Unreal Engine is capable of doing with very little effort. 

After importing the character into Unreal, I noticed there were a few "glitches" in the character model. So I went back to Blender, fixed them, and then reimported the specific asset I wanted to fix. What impressed me was that the workflow was very smooth.

The biggest shock to me has been working with Unreal's Blueprint system. I followed some tutorial videos on setting up the animation states for my character, and it was honestly incredibly easy and quick to set up (relatively technologically complex) animation blending between running, idling, and jumping. I believe that the Blueprint system is preferred for things like that for the ease of use in swapping different animations out at run time and so forth, but I'm still too much of a noob to know anything. Again, what impressed me was the ease in working with the system.

Also, Blender has been much simpler to work with recently. I have a decent workflow in place, and I'm sticking to very simple models to keep things moving quickly. I'm now very comfortable with the idea that I'll have to model and animate dozens of different enemies and level assets. I'm still going to keep things simple and take advantage of old school techniques for reusing assets, but it's nice knowing that I won't be completely overwhelmed when I start to make an even larger game with more detailed models and even more characters.

So far, the thing I miss the most is writing code. I used to spin up Android Studio and spend hours writing code to work on games. It was a slow process to do anything. I spent about a month on getting the shape renderer and animation framework to work as I wanted it to, but I had a lot of fun doing so. With Unreal and Blender, I was able to - in 1 week - get almost the exact same character as Robo from my 2D Robot Ops game fully modeled, animated, and controllable in 3D. Even crazier, it took me a week to do that while learning how to do it, which means that an experienced Unreal/Blender developer would probably spend an afternoon or less getting to the point where I am now.

The ease of development and the change of environment leaves me with mixed feelings. For one thing, I was really hoping to spend more time in C++ for this project. After starting with the Blueprint model, though, I just don't see how I can make the switch for this game; the game is so simple that I think the Blueprint model will be more than enough to handle every bit of logic I need for it. Sticking with Blueprints will just make me quicker.

Another mixed feeling I have is the realization that I could have cut development time in half for the entirety of last year while I was working on Robot Ops. I'm honestly considering doing a Robot Ops 2.5D sequel to the first game, just because of how easy it would be to build with Blender and Unreal.

The ease of development has made me realize that I'll be able to spend time developing the actual game. I will not have to build any tool to make the game I want; everything is already there. That's probably the greatest thing about Unreal Engine (and other engines) for game developers. Using an engine, the focus can be less on software development and more on game development. Of course, this saddens me a bit; I really love writing code. But I also like making games, and I'm sure that there is a lot more code to write ahead of me than I'm expecting.
Leave a Comment