The Silver Lining Progress Report

by on Jan.31, 2019, under featured, tsl

First things first: No, this isn’t a release date announcement. It’s a progress report!

The first of a few we’re planning for the next few weeks, in fact. It’s been a while since we talked about the state of The Silver Lining, and we’d like to let you know some of the things going on behind-the-scenes.

One of the reasons for the slow updates is that much of the work that’s been done in the last year isn’t the very glamorous kind. Much of game development is programming (over and over and over), fixing bugs (over and over and over), exporting animations, rigging, scripting, tweaking, over and over and…you get the idea. In fact, that stuff is the meat of game development, if I’m quite honest.

This year, our TSL work has been focused on the non-glamorous nuts and bolts. In particular, we’ve been making pipeline improvements. Meaning that we’ve been improving our own workflow and the tools we use in order to make said nuts and bolts work better and work for us. For example:

  • Cloth simulation
  • Facial capture and lip-syncing
  • Online VR mocap systems
  • Puzzle tweaking

What each of these entails I’ll go into when we cover them in more detail, but at a high level, here’s the scoop.

Cloth simulation. This one even puzzled me for a while. Sure, every game dev prides themselves on pushing the envelope, honing their craft, making each game better than the last. But we don’t want the look of Ep 5 to stray so far from the previous episodes that it’s jarring, and we’re not trying to match AAA quality. But if you’ve been playing TSL, you may have noticed it’s a little fixated on these people who love wearing cloaks! And what do cloaks do when you move? They also move. And what’s a pain in the butt to animate in every single animation you create for a person in a cloak? The animation of their cloak! You know what helps with that? A tool that handles the animation of the cloth for you! (All of this is also a roundabout way of saying, you’re going to see a lot more of people in the Black and Silver Cloak Societies in Episode 5 than you did previously!)

Facial capture and lip syncing. Much like the cloth simulator, this is a tool that, once implemented, makes animating a lot easier. We have tools that already form a character’s mouth roughly to the sounds of the words, but we still need to plan their expression as they speak, and program those expressions in. Facial capture lets you skip a lot of that by recording the expression of the actor when they read the lines. Now, an added challenge for us is that we recorded all these lines years ago, and without filming the actors, so to get this to work, someone needs to re-read the lines so their face can be captured. And you know how wordy we are– when else has a game had to offer a “Short Narrations” option? So this re-recording takes a while, especially when you’ve got busy schedules and different time zones to work around.

Online VR mocap systems. Full motion capture is just like the facial capture, but it takes a lot more people and a lot more space. Here we’ve got the added challenges of needing physical space to move in, cameras to capture the movement from many angles, and the fact that the person with the cameras and space and the actors lives in a totally different time zone (country, even!) from the person directing the acting. And Phoenix, as you know, is an online company without a physical office. So how do we solve this? Create a virtual space where everyone can be at the same time! With the use of VR tools and headsets, it’s like everyone is in the same place and watching one another live, even seeing the animation on the character model practically live! Instant tweaking and adjustments become possible! There’s still that scheduling issue, of course, and the sheer volume of animations to get through.

Puzzle tweaking. The puzzles of Episode 5 were designed a while ago, but often when you’re working through implementing puzzles in-game, you find you need to make adjustments. We’ve held some brainstorming sessions to address this exact issue, finding ways to keep the puzzles interesting and true to the world and gameplay we’ve established, but also work within the game engine and the ‘physical’ world of the game. I don’t want to give away too much here, since we’re talking about the meat and potatoes of the game, but I will say a forced perspective of distance posed a problem when a distant object that needed to be interacted with proved to be further away due to the scene setup than we thought it was thanks to looking much closer due to the angle of the camera. Moving the camera didn’t work as a solution, because everything else was framed up where we needed it, so we had to find a way to fit fewer objects in a smaller space that was in fact a much larger space. Basically, game design is like a TARDIS, sometimes!

In the coming weeks, we’ll gather and share some more of the ‘flashy’ examples of some of these things and go into some more details. Thanks as always for coming along on this journey with us!


Comments are closed.