Making TESSELATE - Part 2

02//kaleidoscope



The level was done.

I looked at this level over and over and it was so amazing to me every time I pressed play. Something that actually worked, yknow. As a programmer, wow. I could appreciate this level all day. The mechanics are so solid, the objective is there. To me, this level was done.

But the look wasn't very visually appealing.
________________________________________________

Concept art for the stage was blocky and very "clean". I wanted the player to feel like they were exploring an abandoned space station. They would be picking up these weird cube-objects and discover the alien technology that was left behind. Set pieces would consist of "atom"-like objects, using the gravity script I made. Scattered "altars" and other large structures would dot the landscape for the player to explore.

Low-poly mountains really helped us visualize how we wanted the scenery to look like. Clean and smooth flat faces. Multiple light sources also help bring out the surrealistic look, ambient lighting will be used to our advantage here.

I even tried modeling. Imagining an object in 3D space and having to pinpoint the exact location of every vertex is excruciatingly difficult, and I am so in awe of how dedicated 3D artists are.

________________________________________________

I decided to take a page out of Bethesda's book of level creation tricks. I modeled a bunch of generic "rock" models and placed them around the level overlapping each other to simulate "random" rock formations.

As you can see in the above screenshot, we used a license-free Space Skybox pack that included I think 6 skyboxes. Also, in the editor view you can see how I just overlapped rocks to act as the walls of the level.

The level looked really cool already. Almost exactly like how I envisioned it in my head.

We were using Unity 4 at the time, and colored lights blended into each other. I wanted to use the lights to add color to the level. Give it a real "Hotline Bling" feel.

Adding bright vibrant colors made the level feel very different. I experimented with having a whole ton of lights, to help the player associate certain areas with specific colors. I felt like something was missing though.

Erickson suggested we try Cel-shading, which is a cartoon-like style where edges are drawn with outlines. Similar to Borderlands or Sly Cooper. The above image is the visualization of "Edge Detection", which drew lines on sharp edges detected by the camera.

Sample of how colors and shadows looked with edge detection turned on. I took this screenshot because it reminded me so much of Borderlands, I had to show the team in the chat. Haha

Edge Detection sensitivity turned up, the script draws lines based on the change in geometry detected by the camera.

This is a nice screenshot taken right after a trial run of the level. It was ready to be played and if you knew exactly what to do, you could finish the level in 12 minutes.

Another view of Editor Mode.

__________________________________

At this point, I'd already shown the level to a few people and their feedback was.. not what I expected haha. It seems that even if you are so convinced that an idea looks great to you, it might not be that way for other people.

So, the level looked "incomplete" because the shapes and formations did not have textures on them. And while my initial idea was the they really -wouldn't- have textures, I believed it was a good idea to follow the feedback and adjust my vision.

The problem was that the rocks that I had originally created to be used.. did not accept textures gracefully. I literally know nothing about texturing and I had to scrap the whole shiny white rocks idea. :(

The way around this was to recreate the entire level with Cubes, using the same technique. Using big cubes and small cubes to simulate rock formations.

____________________________________

So here, you can see that I applied a Normal Map to some of the objects in the level.

The tri-cube tower is one of the set pieces that I used to help the player navigate through the level. We actually used this screenshot for our final poster! In a way, simulating the player's experience running into these structures. A hint to what the game is like.

Our poster was gigantic. That was fun. A sneak peak at Incendium VI, which I will talk about in a later post.

Another Cube Formation that I made to kind of give the level a little life. The cubes floated using a sin-wave function and looked like they were moving in a wave pattern. Really fun.

This is a screenshot of a somewhat completed Level 4. The team sort of settled on the brown sandy look because of the mountain skybox we found. It sort of gave it an epic looking feel.

One of the set pieces I'm really proud of, the hypercube. You can also see me editing the level boundaries in the editor view on the right screen. If the player touched them, it would cause the player to respawn.

The hypercube. Sadly we were unable to use it in the final game, because I wanted to expand its usage. Perhaps in the next game...

______________________________

So that ends this post, which actually had a lot of screenshots!! Fun times! I talked about how we settled the aesthetic and look of the level. Thinking back, it really is a lot of experimentation. You should try all kinds of looks before sticking to one that you like.

At this point, it was already the midterms and we presented the game as it was. Mountainous background and sandy/blocky. In the next post, I'll talk about the midterm defense!

______________________________

It's been such a looooooong time since I've updated my blog.. Lots of things happened, I graduated! I spent some time bumming, I got a job doing research for app dev. Cool cool stuff that took SO much of my time. But now I want to continue this blog to really get into the creative process. Yknow... before I forget hahaha.

It's actually been an entire year since TESSELATE and I am going through so much nostalgia right now writing this blog post. I loved my thesis project, so many great memories, so many amazing lessons. Time to work on the sequel~