We're pretty happy with our entry for LD43 this year, even though we didn't get to build the ending we wanted and had to scrape a bunch of content, this might be our least buggy entry. @linear and I have been doing Ludum Dare for a while, and LD43 was our... 7th Ludum Dare, I believe.

Check it out! Will you save the world? We'd love to hear your honest feedback about how our puzzles played out and how the game plays out!

https://ldjam.com/events/ludum-dare/43/scapegoat

A Different Art Style This Year

That said, I wanted to share what we did for an art style. In our past LDs, we took out all of the techniques we could think of... Parallax Occlusion Mapping? Check. Tesselation? Check. Complex shadows with custom shadow casters? Check. Detail Mapping? Check.

That said, it's hard to get that to look good during a Ludum, so we decided to go for an art style that is different for our team. We decided to analyze The Legend of Zelda: Twilight Princess to see how they do things. Thus, before LD even started, we produced an art guide for ourselves so that we could make sure we were on the same page as we build different assets and prefabs.

Scapegoat Art Guide - How to Mimic TP

Our attempt

With only two people actually working in Unity proper, and myself doing most of the code and @linear doing most of the art, we couldn't get the EXACT results we wanted, but we think we did a decent job. We chose to sacrifice normal maps for the first time, which is pain when you've been generating normals for years, and relied wholly on Ambient Occlusion.

Considering how we didn't use any Unity Terrain but all custom geometry, the ground blending turned out okay, but cliff blending took much longer to get a decent result:

We tried to make sure that things such as levers and the goat were visible against the intentionally stretched background, but it turns out we didn't get the effect we wanted as too many people didn't see levers or keys scattered through the levels.

In order to get the desired blending, we ended up creating a number of blending masks that we used even for detailing things:

Different masks served different purposes, but a lot of the ground blending was done using basic photoshop filters applied to black and white clouds.

In order to give our space a nice vibe, we used a tree shadow spot light with fake tree branches and leaves, to imitate the effect of actual shadows from above (which didn't exist).

Finally, our cavern sequences turned out to be really hard to see the goat in because of the purple-ish light we had emitted from the crystals and the goat's white fur. In the picture below, the goat is hardly visible, when in fact he ought to be higher detail density than the rest of the background. Alas, these are things we only noticed when watching others play our game after LD!

And some things we made too detailed. For example, these mushrooms were just little decorators, but people kept thinking they were important. This is why consistency is important in detail level.

Thoughts

All in all, there were definitely things we could have done a lot better with our graphics, but I think we found a nice art style that doesn't let us fall into the uncanny valley and is still doable in a short amount of time. The lesson we learned is that sacrificing graphical fidelity doesn't ruin the graphical aesthetic. Even if you have some seams and places where objects intersect a bit, it's not the end of the world for the atmosphere and play of the game. We're pretty happy with how things turned out aesthetically, but we are looking forward to improving for next time now that we've got this technique figured out.

Thanks for reading!

Check out our game! and please let us know what you think!