Friday, 10 February 2017

Technobabylon 2, Part 1: An Evolutionary Process

In a few months, it will have been two years since Technobabylon was released. Part of me, therefore, is panicking that I've not managed to release another game to follow it up yet - after all, that's a pretty hefty amount of time - but everyone can be assured that I'm not being completely bone-idle.

The first game was built with AGS, and I'd decided to move on to Unity for the next project. The downside of that was having to actually learn how to use the new system. On the plus side, it works a lot better by default with my favoured artistic medium of Blender. And, with eighteen months of messing around, I've managed to learn a few things about it.

To prove that things are actually getting done, here's an overview of the various Technobabylon-related things that have been produced. Granted, a "sequel" has taken several different shapes since April 2015, but with every shift things get better.

First Shot: Early Experiments

Making another game was, unsurprisingly, a foregone conclusion. People liked playing Technobabylon, and I really enjoyed making it. Of course there'd be another game! And with the ability to make use of 3D as a medium rather than an art reference opened a lot of doors. To make nice-looking characters though, I needed nice-looking references. The fanart of Amanda Virtudes was a good place to start from, so we produced some turnarounds for two of the main characters of the game. From an "essentials" perspective, it was a good place to begin from - Regis would make a good basis for a "default male" mesh/skeleton, and Lao for the "default female".

Planning for the Dr Regis Reference

Planning for the Dr Lao Reference

Planning for the Latha/Mandala Reference

With a bit more outsourcing, Ivan Ulyanov discovered a talent for texturing as we ended up with a nice-looking Doctor Lao. With these polygonal protagonists now on-hand, experiments could begin in the third dimension, to see what Unity could do - for example, importing motion-captured animations from the Carnegie Mellon database.


In Technobabylon, much of the scenery was designed in 3D using Blender, then used by Ben Chandler to create appropriate-looking pixel-art backgrounds, a system that served us well. My first (and laziest) go at making scenery in 3D involved simply pasting the 2D scene on top of the existing 3D model ("unwrap > project from view", if you're interested). The results were...better than expected. Obviously, any turning of the camera would create funny warping effects, and areas behind obstacles that hadn't been painted would also come out a little peculiarly.

The original 3D reference for Technobabylon's Aerostat hallway



Also, the low resolution didn't serve the 3D well, with overly-crisp 3D characters against a 320x200 background. However, it was an interesting place to begin from! A couple of attempts were made at higher-resolution paste-on textures, and although interesting, it didn't take advantage of the opportunities that Blender and Unity provided.

One of those advantages, I must say, was Adventure Creator, an addon for Unity. It essentially provides a framework within Unity for designing adventure games in a manner that is familiar to those who've used AGS. It's got differences here and there, and you've still got to get used to some of Unity's own aspects (for example, the way it handles GUIs), but for making an adventure, it'll cut so much time out of building the framework for it - on its own, Unity is essentially a blank slate, AC gives it a kickstart in the right direction. Plus, its designer, Chris Burton, is extraordinarily helpful - any daft questions you've got, you'll find them answered swiftly on the forums (or a polite pointer to where in the manual to find the info). So yes, another good argument for using it.

Another long-term goal I have is to ensure that future games will work on mobile platforms. With AGS, I can't even imagine where to begin with such a task, but in Unity it simply became a matter of ticking an "Android" box.


So, in the summer of 2015, the experiments had begun, and I was starting to get the hang of how one drives Unity.

Next episode: ISOBABYLON

No comments:

Post a Comment