Date: February 16th – February 23rd, 2017
Engine + Platform: Unreal Engine 4, PC
Description and Details
Welcome to the Castle Valley River, a level made in UE4 where the player is contained inside of a small valley with a castle structure some distance above them.
I created this level without any real major gameplay and I did so on-purpose. I strive to be a Level Designer and besides simply planning out and creating/scripting gameplay, a big part of a Level Designer’s job is to simply create the environment using assets that they are given by the art team.
I created this level without gameplay as a way to specifically test and improve upon my blockout abilities using assets provided by Unreal Engine that I did not make myself.
That might be my intent for this project, but I also designed the level with gameplay in mind, which can and will (hopefully) be implemented at a later date. I very much enjoy Deathmatch and Capture the Flag (CTF) game modes in competitive games, so I went about picking the level’s smaller size for two reasons: to keep the project manageable and to force a different kind of gameplay design for CTF games. We’ll get into that in a little bit.
Research and Planning
I’ve designed a few levels in my time, but the biggest pattern that I found among most of my levels is that they all take place completely or mostly inside. Upon figuring this out, I decided that it would benefit me most to design a level that took place completely outside to test myself and add variety to my portfolio and my skillset. For some reason, when I thought about outside, I thought castle over and over and that’s what really pushed me to start researching things like castle courtyards and valleys. Check out some of the reference images I used:
The big thing that I noticed in a lot of the reference images that I found, like the ones above, was that the size of each valley (or courtyard) seemed to scale with the size of the castles associated with them. With that, I started visualizing my level as a small, enclosed valley with a “small” castle looking down upon it and the player.
My pre-production phase for this project was actually pretty short, but I did a decent amount. First off, I made a very rough sketch of my level based off of what I though I could make it look like. The sketch can be seen below:
Rough (really rough), I know, but it got the job done. Following this, because of the fact that I would not be creating my own assets for this project, I decided to look through the free assets that I could obtain from the Unreal Marketplace as well as some of the paid asset packs that I already owned. I based the assets that I really searched for off of both my rough sketch and the reference images that I used and I was quite happy with what I found, happy enough to move forward with this project idea.
During the production phase of this project, I followed the Epic Games level design model fairly closely, though not exactly to the letter. This model included doing several passes of the level such as a prototype pass, meshing pass, lighting pass and polish pass. I didn’t follow this method to the letter, but I followed it somewhat loosely. I started my design with an incredibly rough prototype pass using just primitives and BSP blocks in the engine, following which I moved into the meshing pass where I placed down all of the assets that I would be using for the level. This second pass took the longest of all, as getting the level to look good aesthetically was a bit difficult throughout. Finally, I combined the lighting and polish passes into what I called my finishing pass. The slideshow below shows off screenshots of the level in each of my three passes.
Before I start, it’s important to remember that this level has no gameplay in its current state and that I created it specifically to test my blockout abilities using assets that I myself did not create. Knowing that, I can absolutely say that I learned quite a bit during this project’s development.
First off, the level was designed, as I mentioned, using Unreal Engine 4 and using it extensively in this manner to create a level meant that I learned a lot more about what makes the engine tick. This specifically stood out when it came to lighting my level. You can see the final, lit version in the finishing pass photo above but even with that, the lighting is nowhere near perfect. I actually spent about 2 hours just trying to get the angle of the sun correct so that the overall environmental lighting would look decent. A big problem that I realized may have caused this long lighting period was the fact that while I did find environmental reference, I didn’t search for lighting references based off of my chosen environment. In the end, I think that going that extra mile and looking up lighting references for projects before moving into lighting passes or even the production phase will do me nothing but good.
Don’t use a lot of grass when designing levels like this, at least not if you’re trying to make something playable. If I must reveal my secret, there are a little over 70,000 grass instances on this map and it’s made even worse when we look at how small the level actually is. True, I was not designing this level to have gameplay incorporated but I did learn quite a lot about the foliage tool and the limits of its use along the way. Plus, the amount of grass instances also made building lighting take about 25 minutes every time I did it, which really wasn’t fun.
Going back to earlier, I mentioned that a big inspiration for this project’s setting was the fact that most levels I’ve designed have been inside. I’m quite proud of how the level turned out, especially given that it was my first time making one completely outside. Something I did learn, however, was that making a level outside like this is, in my opinion, harder than designing an interior level. Dealing with the rest of the environment that isn’t enclosed becomes a much bigger job, and when I don’t have the proper texture sets to cover the small mountains that surround my level, they end up being gray with the default landscape texture. Just as well, without specifically creating something myself to serve the purpose, I wasn’t able to make the water look as realistic as I could. Knowing that has only made me want to look into creating more realistic and even swimmable water in future outdoor projects.
Overall, I certainly learned a lot throughout this project’s development. While I ran into a handful of obstacles and not everything turned out quite the way I wanted it to, I’d still say that I’m quite proud of the final product and I came out as a better level designer because of it.