If there’s one thing I’ve learned developing Terror High, it’s that writing regular devlogs is not my strong suit, and I’m much more interested in environmental design than nuts-and-bolts game mechanics…
Okay, the two things I’ve learned developing Terror High are writing regular devlogs and working on nuts-and-bolts mechanics are not my strong suit, and I’m also easily distracted by storytelling, lore, and world building…
The three things I’ve learned…amongst the things I’ve learned…
I’ll come in again.
It’s been 1010 days since my last devlog, Terrorteers. That’s, like, 36 generations of houseflies! One-and-a-half-times longer than the average wait between seasons of The Venture Bros.!* 9.3% of Saturn’s sidereal orbit!
Point is, it’s been a while. That’s partially because I work full time and in the limited time I have for Terror High, I’d rather work on the game than talk about working on the game. It’s also partially because Twitter’s a little easier way to update y’all, because even though it drives all its users to madness, one can simply post snippets and screenshots here and there in between reading some of the absolute worst Takes™ in recorded history. A devlog, on the other hand, requires multiple, whole sentences. Paragraphs, even! Ain’t no one got time for that.
And the longer I go without writing this devlog, the longer it takes to write. It’s like leaving a carton of milk in the fridge. You’re pretty sure it’s right on the edge, but you don’t want to throw it out just yet, ‘cause maybe it’s still good, so you leave it. And then you remember a week later and you’re pretty sure it’s definitely gone off since then, but now it’s even more of a hassle, plus you’re kind of ashamed you didn’t take care of it sooner, so you put it off even longer, until eventually it’s just become A Science Experiment taking up fridge space. (Or maybe you don’t do this, dear reader, and I’m going to find out this is yet another symptom of undiagnosed-probably-ADHD.)
Thus, my gentle Terrorteers, you sit here waiting for updates, updates that never come. So here I am. I am here. Writing an update. To update you, you understand. So you’re up to date.
⁂
A lot has changed since I started rebuilding Terror High in the Unity engine.
After the move to Unity, I made the decision to switch from sprite-based characters to 3D modeled characters, owing to the sheer workload involved in sprite animation. This allows me to use those previously covered motion capture animations almost immediately, rather than after a painstaking rotoscoping process. (That said, I have also been looking into other motion capture solutions, as relying on only preexisting mocaps is proving somewhat limiting…)
It also allows me to iterate character designs as needed, rather than feeling beholden to an earlier design because I spent hours upon hours pixeling a li’l guy into existence. However, this also meant I had to step up my 3D modeling game, which luckily hasn’t been as steep of a learning curve as I once feared it would be. (Though I will admit to seeking the aide of some assistance in the form of Make Human, as my modeling skills are still generally stronger in what Adam Savage would describe as “hard-edged model making”.)
Something else I’ve had to deal with in the switch to Unity and 3D is the rendering of the game world. The Enter the Gungeon method of scaling and rotating the world by sqrt(2) and 45° respectively (as detailed in the last devlog) worked for a time, but it had a lot of knock-on effects with physics, character rigging, pathfinding, etc. It just became too much of a headache each time a new wrinkle popped up, so I’ve opted for a much simpler method of simply scaling the camera’s projection matrix vertically, et voila, Robert’s your parent’s sibling, we get the same results without causing the Jenga tower to topple. (I am having a bit of trouble migrating the decal system to the new projection, but my grandma always said you can’t make an omelette without breaking a few decal projections.)
I’m also inclined by this shift into the third dimension to treat the town of Pike’s Landing and the surrounding geography more naturally and organically, rather than a simple tilemap. This involved even more “location scouting” for suitable geographic locations to base the game world on, as I find that real-world reference is always going to make something feel more believable than completely winging it.
Pike’s Landing is a seaside town in Oregon in the fictional Calhoun County, so I focused my search on the coastal regions of the Pacific Northwest. I really loved the look of the coastline around Port Orford, OR and, combined with some elements of Tillamook Bay, I cobbled together a heightmap in Photoshop that I could import into Blender. (If you’re interested in doing something similar, both terrain.party and the Tangram Heightmapper are options, though it’s been a while since I went down that rabbit hole and I recall both had their pros and cons.)
And, of course, “location scouting” continues in the form of building up the architecture of the town, as well.
This might seem like the wrong thing to prioritize, and I’m not even going to disagree with anyone on that, but it’s really important to me that Pike’s Landing be a place the player can inhabit and explore, because frankly that’s one of the things I enjoy and want from the games I play. (The news that we’ll be able to explore Bright Falls as Saga Anderson in Alan Wake II has me positively giddy.)
Furthermore, going back to one of my theses for Terror High, I think the contrast between our everyday world and the creeping horrors just beyond the edges of our perception is what makes a story frightening. For me, that juxtaposition is why stories like Pet Semetery or Midnight Mass work, and it’s something I want to bring to my own world.
Also I feel like I was robbed as a child for not being able to bike around town solving mysteries so I am putting right what once went wrong, gosh darn it.
I’ve also been doing all sorts of other work on the game, ranging from background props to fleshing out the story to experimenting with various ways of implementing the NPC and enemy AI, like finite state machines to behavior trees, but those are topics for another day (or year). I am but one man, and gamedev takes time. So much time…
I’ll try my best to write more regularly, but until next time, Terrorteers – stay spooky.
* Yes, I did the math. I didn’t include the specials but I did count the two halves of season 4 as two different wait times.
Terror High 
